davinci-resolve-mcp

Configure MCP conversational defaults and setup preferences. Actions: schema() -> {defaults, actions} get_defaults() -> {defaults} set_defaults(defaults?|media_analysis?|updates?|dry_run?) -> {defaults, changes} clear_defaults(keys?, dry_run?) -> {defaults, cleared} Current defaults: media_analysis.*: analysis, metadata, marker, reporting, and workflow defaults updates.*: MCP update policy, interval, and snooze defaults

App-level DaVinci Resolve operations. Actions: launch(headless?) -> {success, message, running, headless, guidance} — Launch DaVinci Resolve if not running. Call this FIRST if any tool returns a 'Not connected' error. headless=true starts it with no UI (-nogui): capability-identical to the GUI and immune to the modal dialogs that can block a script indefinitely. Errors rather than starting a second instance if one is already running in the other mode. runtime_mode() -> {running, headless, instances, guidance} — Is Resolve up, and does it have a UI? Answers with no connection needed. headless is null when it cannot be determined — do NOT read that as false. There is no way to tell from the scripting API itself; this reads the process argv. Consult before any LoadProject/CloseProject: with a UI, an unsaved outgoing project raises a dialog no script can dismiss. get_version() -> {product, version, version_string, mcp: {version, update, update_decision}} — mcp.update_decision piggybacks the cached update check: if its action is "notify" or "prompt", mention the available MCP update to the user ONCE per session (do not nag, do not auto-apply; updates are applied via install.py or the control panel). mcp_update_status(force_check?) -> {version, update, decision} set_mcp_update_policy(mode) -> {success, version, update, decision} ignore_mcp_update() -> {success, version, update, decision} snooze_mcp_update(hours?) -> {success, version, update, decision} clear_mcp_update_preferences() -> {success, version, update, decision} api_truth(query?, resolve_version?) -> {verified_on, count, facts, live_version?} — look up behaviorally-verified facts about quirky/unreliable Resolve API behavior (no connection needed; never connects). When a build is already connected or resolve_version is passed, each fact carries a version_context saying whether it was measured on an older, newer or identical build — the API changes per PATCH release, so a fact from another build is a prior, not a finding. check_version_support(symbol?, resolve_version?) -> {live_version, support|unavailable_on_this_build} — does THIS build have that method? Answers "unknown" unless a gate is recorded, because most of the API has never been version-bisected and a false "available" is how an agent insists a missing method exists. Omit symbol for every recorded gate this build does not clear. verification_stats() -> {stats} — readback-verification tally (verified/contradicted/unverified) since server start (no connection needed). job_status(job_id) -> {id, label, status, result?, error?, started_at, ended_at} — poll a background job started by a long op run with background=True (no connection needed). status is running, done, or error. list_jobs() -> {jobs} — compact status of every known background job (no connection needed). get_page() -> {page} open_page(page) -> {success} — page: edit, cut, color, fusion, fairlight, deliver get_keyframe_mode() -> {mode} set_keyframe_mode(mode) -> {success} quit() -> {success} get_fairlight_presets() -> {presets} set_high_priority() -> {success} disable_background_tasks_for_current_session() -> {success} — Resolve 21+ list_user_preferences_presets() -> {presets} — Resolve 21.0.4+ save_user_preferences_preset(name) -> {success} — Resolve 21.0.4+ load_user_preferences_preset(name) -> {success} — Resolve 21.0.4+. SESSION-WIDE: swaps the user's global Resolve preferences, not a project setting. Only call when the user asked for the switch. delete_user_preferences_preset(name) -> {success} — Resolve 21.0.4+ import_user_preferences_preset(path, name?) -> {success} — Resolve 21.0.4+. The imported preset is NOT auto-loaded; follow with load_user_preferences_preset to activate it. Omitting name is safe (reported on 21.0.4.5: single-arg returns True, not a TypeError) and the preset is then named after the file, so pass name when the preset should be called something the filename does not say. export_user_preferences_preset(name, path) -> {success} — Resolve 21.0.4+ open_control_panel(port?, open_browser?, force_restart?) -> {success, url, pid, port, status} — Launches the analysis control panel (src/analysis_dashboard.py) as a background process. Idempotent: returns the existing URL if already running. Binds 127.0.0.1 ONLY — a `host` other than loopback is refused, there is no override. The returned url carries a per-launch bearer token in its fragment (#token=…); the panel refuses every request without it, so give the user that exact URL, not a bare http://127.0.0.1:8765. control_panel_status() -> {running, pid, port, url} close_control_panel() -> {success, was_running} save_state() -> {state_token, page, current_timeline_id, current_timecode, selected_clip_ids} — Captures the current Resolve UI state so it can be restored after a preview. restore_state(state_token) -> {success, restored: {...}} — Returns Resolve to a previously-saved state.

Manage DaVinci Resolve UI layout presets. Actions: list() -> {presets} — Resolve 21.0.4+; saved layout preset names save(name) -> {success} load(name) -> {success} update(name) -> {success} export(name, path) -> {success} import_preset(path, name?) -> {success} delete(name) -> {success}

Timeline operations: tracks, clips, import/export, generators, titles. PREFER probe_* / *_capabilities / *_boundary_report / *_checked / dry_run variants where they exist. Raw mutators are kept for advanced callers but bypass guardrails. DESTRUCTIVE actions auto-archive the current timeline (C6); pass strict=true for refuse-on-archive-failure behavior on the catastrophic ops. To thread several destructive calls into ONE archived predecessor instead of N, wrap them in timeline_versioning begin_run/end_run. Frame numbers are TIMELINE/record frames (position on the timeline) unless an action says SOURCE. Source frames are positions within a media-pool clip's own media: create_variant_from_ranges takes SOURCE start_frame/end_frame; extract_source_frame_ranges and source_range_report return SOURCE ranges. A SOURCE frame is counted in the MEDIA's own frame rate, not the timeline's: an AUDIO item's source_start/source_end read back in the file's rate. A WAV has no intrinsic rate and freezes the PROJECT's rate at import, so it differs from the timeline whenever the project moved afterwards — read source_fps, never assume 24, and converting at the timeline rate is silently wrong by minutes (resolve_control api_truth "GetSourceStartFrame"). Actions: list() -> {timelines} get_current() -> {name, id, start_frame, end_frame, start_timecode} set_current(index|id|name) -> {success} — id/name are stable across archives; index is 1-based get_name() -> {name} set_name(name) -> {success} get_start_frame() -> {frame} get_end_frame() -> {frame} get_start_timecode() -> {timecode} set_start_timecode(timecode) -> {success} get_track_count(track_type) -> {count} — video, audio, subtitle add_track(track_type, options?) -> {success} — options dict (newTrackOptions per docs line 327): {audio_type, index}. audio_type: 'mono', 'stereo', '5.1', '7.1', 'adaptive1'..'adaptive36' for audio. index: 1-based slot; appended if omitted/out of bounds. delete_track(track_type, index) -> {success} CATASTROPHIC. Deletes a track and all items on it. Strict mode auto-applies. get_track_sub_type(track_type, index) -> {sub_type} set_track_enable(track_type, index, enabled) -> {success} get_track_enabled(track_type, index) -> {enabled} set_track_lock(track_type, index, locked) -> {success} get_track_locked(track_type, index) -> {locked} get_track_name(track_type, index) -> {name} set_track_name(track_type, index, name) -> {success} get_items(track_type, index|track_index) -> {items} — track_type: video|audio|subtitle clip_where(filters|track_type?, track_index?, name_contains?, duration_lt?, duration_gt?) -> {clips, match_count, total_clips} Find clips on the current timeline matching named filters (AND). Pass filters inline or as a `filters` dict. Live filters: track_type, track_index, name_contains, duration_lt, duration_gt (frames). No clip enumeration needed. delete_clips(clip_ids, ripple?) -> {success} — clip_ids: list of unique IDs DESTRUCTIVE. ripple=True is CATASTROPHIC (closes the gap; cannot be selectively undone). set_clips_linked(clip_ids, linked) -> {success} duplicate(name?) -> {success, name} duplicate_clips(clip_ids?, selected?, target_track_index?, track_offset?, placement?, record_frame?, record_frame_offset?, copy_properties?, include_linked?) -> {results, count} — Video clips only. Re-places the same MediaPool media with the same source trim on the current timeline (like Alt-drag) via AppendToTimeline. clip_ids: timeline item unique IDs from get_items / get_current_video_item; selected=True uses Resolve's selected/current item when available. target_track_index overrides track_offset; placement supports same_time, offset, at_playhead, track_above, after_source, and next_gap. copy_properties may include transform, crop, composite, audio, retime, clip_color, markers, flags, enabled, cache, voice_isolation, fusion, grades, takes, and transitions (reported unsupported by Resolve API). include_linked=True duplicates linked audio and restores link state. # example: action_help(name='<action_name>') copy_clips(...) -> {results, count} — alias for duplicate_clips. move_clips(...) -> {results, count, deleted_sources} — duplicate, then delete successfully duplicated sources. copy_range/duplicate_range(start_frame, end_frame, record_frame, ...) -> {results, count} overwrite_range(start_frame, end_frame, record_frame, ...) -> {results, count} lift_range(start_frame, end_frame, allow_partial_item_delete?, ripple?) -> {success, deleted} DESTRUCTIVE. Deletes timeline items whose record-frame range overlaps [start,end]. ripple=True closes the gap left by those deleted items (shifts later clips left). It does NOT close a pre-existing empty gap: if no item overlaps the range, deleted=0 and nothing moves. (frames here are TIMELINE/record frames.) story_spine_report() -> {beats, track_summaries, source_ranges, audio_spine} create_variant_from_ranges(name, ranges, markers?, cdl?, dry_run?) -> {success, id, items} ranges[] take track_type? (video|audio) and track_index? (1-based, within the track_type, default 1); missing tracks are added, so V2/V3 multicam angles survive. # example: action_help(name='<action_name>') bulk_set_item_properties(ops, dry_run?, readback?) -> {results, op_count} # example: action_help(name='<action_name>') apply_look_to_items(target_ids, cdl?|copy_from_item_id?, dry_run?) -> {success} # example: action_help(name='<action_name>') thumbnail_contact_sheet(frames?|max_samples?, analysis_root?) -> {path, samples} frames are relative to the timeline start (frame 0 = first frame), like marker frameIds. marker_thumbnail_review(max_samples?, analysis_root?) -> {path, samples, review_guidance} edit_kernel_capabilities() -> {supported, partially_supported, unsupported} probe_edit_kernel_item(clip_ids? selected? timeline_item?) -> {items, count} title_property_scan(clip_id|timeline_item_id|timeline_item) -> {properties, text_key_candidates, fusion_comp_count, ...} — Undocumented generator/Text+ GetProperty map on an item (keys are not in public API docs). set_title_text(clip_id|..., text, property_key?, as_styled_xml?, try_plain_first?, try_heuristic_keys?, readback?) -> {success, property_key?, attempts} — SetProperty on a heuristic or explicit key; tries plain string then minimal styled XML unless as_styled_xml=True. # example: action_help(name='<action_name>') bulk_set_title_text(ops, ...) -> {results, op_count} — list of set_title_text payloads (same params per op). create_compound_clip(clip_ids, info?) -> {success} create_fusion_clip(clip_ids) -> {success} import_into_timeline(path, options?) -> {success} UNSAFE. No path sandboxing. Prefer import_timeline_checked. export(path, type, subtype?, background?) -> {success | job_id} — type: AAF, EDL, FCPXML, etc. UNSAFE. No path sandboxing. Prefer export_timeline_checked. get_setting(name?) -> {settings} set_setting(name, value) -> {success, known_limitation?} A refusal carries the api_truth entry for that key when one exists. insert_generator(name) -> {success} insert_fusion_generator(name) -> {success} insert_fusion_composition() -> {success} insert_ofx_generator(name) -> {success} insert_title(name) -> {success} insert_fusion_title(name) -> {success} get_unique_id() -> {id} get_node_graph() -> {available} get_media_pool_item() -> {name, id} get_transcript(with_timecodes?) -> {text, cue_count, has_subtitles, cues} Read the timeline's subtitle track(s) as transcript text. This is the TIMELINE subtitle transcript that propose_cuts consumes; it is separate from clip-level transcription (produced by transcribe_audio on a folder/clip, read via media_pool_item get_transcription), which does not drive timeline cuts. propose_cuts(cues?, long_pause_frames?) -> {cuts, cut_count, basis_cue_count, pass, note} DRY-RUN. Mechanically detect candidate cuts (filler words, long pauses, repeated lines) from the subtitle transcript. Proposes only; applies nothing. apply_cuts(cuts, dry_run?, confirm_token?, allow_partial_item_delete?) -> {applied, total, skipped, results} skipped lists cuts that were not applicable, each with a reason. Apply a CutList (from propose_cuts) as lift/ripple deletes. DRY-RUN by default; applying is DESTRUCTIVE — confirm-token gated and a timeline version is archived first. Cuts apply latest-first. get_mark_in_out() -> {mark} set_mark_in_out(mark_in, mark_out, type?) -> {success} clear_mark_in_out(type?) -> {success} convert_to_stereo() -> {success} get_items_in_track(track_type, track_index|index) -> {items} — alias of get_items get_voice_isolation_state(track_index) -> {isEnabled, amount} set_voice_isolation_state(track_index, state) -> {success} extract_source_frame_ranges(handles?, gap_max?, skip_extensions?) -> {timeline_name, frame_ranges, occurrences, ...} — Same logic as Pr/extract_timeline_frames.py get_resolve_api_frames: all video clips on the current timeline; clip name = basename of Media Pool File Path when set; skips audio extensions. source_range_final and frame_ranges tuples are inclusive/inclusive endpoints per that script. Default handles=24, gap_max=30. Use handles=0 for gap-only auto handles. conform_capabilities() -> {supported, partially_supported, unsupported, export_aliases} probe_timeline_structure(track_types?, include_markers?, include_clip_properties?) -> {tracks, markers} Each item reports source_start/source_end (SOURCE frames, end EXCLUSIVE) in the MEDIA's frame rate, the source_fps they are counted in, and source_start_seconds/source_end_seconds. Use the reported source_fps — a WAV freezes the PROJECT's rate at import, so it differs from the timeline whenever the project moved afterwards, and dividing by the timeline rate is then wrong by minutes. source_fps is null when the rate could not be read; treat the frames as unitless then, do not assume the timeline's. All four source_* fields are FILE-relative — frame 0 is the head of the media, whatever timecode the camera stamped on it. source_end is the span between Resolve's source second-readers anchored on source_start, so it is a source frame even when the rates differ and even on media with a non-zero start TC (those readers are timecode-absolute; the offset cancels in the span). It falls back to source_start + TIMELINE duration (unit-mixed) only when those readers or the rate are unavailable. detect_gaps_overlaps(track_types?, min_gap?) -> {gaps, overlaps} source_range_report(handles?, merge?) -> {ranges, occurrences} export_timeline_checked(path, format?|type?, subtype?, require_temp_path?, dry_run?, background?) -> {success, path, size | job_id} import_timeline_checked(path, options?, timeline_name?, import_source_clips?, sanitize_media?, require_temp_path?, dry_run?, background?) -> {success, name, id, media, sanitize?, warning? | job_id} Imports an FCP7 xmeml / FCPXML / AAF timeline. AAF (.aaf) is read natively by Resolve — the XML sanitize pass is SKIPPED for it (media relinks via the media pool, not by rewriting the file); a `note` says so. For AAF, passing relink_search_roots triggers a POST-import media-pool relink (RelinkClips against each root) — fuzzy-relink parity by a media-pool mechanism; the `relink` block reports before/after coverage. Premiere .prproj has no Resolve importer — refused with a message pointing to the advanced MCP's offline read + convert_to_interchange bridge (no Premiere needed). `media` reports {total, linked, offline} coverage; `warning` flags offline items. sanitize_media=True (recommended for Premiere/FCP exports) rewrites the XML first — dropping clipitems that reference MISSING media or are generators (slug/solid with no media file), both of which otherwise abort the scripting-API import and leave the whole timeline offline. Returns `sanitize` with the kept/removed breakdown. The original is only read; the sanitized copy goes to a temp dir (no require_temp_path needed when sanitizing). relink_search_roots (str|list) enables fuzzy relink: clips missing at their original path are matched against media under those roots (exact/ext-agnostic/ normalized/reel + IDF true-source) and their pathurl rewritten so they relink instead of being dropped. relink_min_confidence (default 0.7) gates matches; `sanitize.relinked` / `sanitize.ambiguous` report the outcome. verify_visually=True (or passing reference_movie) frame-checks each proposed relink against a reference (reference_movie at the clip's record position, else the offline proxy at the original path) using a brightness-robust SSIM-structure metric; a conflicting relink is VETOED — reverted and reported under `sanitize.flagged` for human review, never silently applied. verify_threshold (default 0.90) sets the structural-match bar. import_from_drp(drpPath, timelineNames?|timelineIndexes?, import_source_clips?, dry_run?) -> {success, selected, imported, available, results} Extract chosen timelines from a .drp (offline zip surgery → temp .drt each) and import each into the running Resolve. Omit the selector to import ALL. Enumerate first with the advanced MCP drt.list_sequences / editorial.list_sequences to get [{id, name, eventCount}] for a "which sequence?" picker. Requires Resolve open. compare_timelines(right_timeline_id?|right_timeline_index?|left_snapshot?, right_snapshot?) -> {match, differences} probe_interchange_roundtrip(format?, output_dir?, cleanup_imported?) -> {success, export, import, comparison} detect_missing_media(sanitized?|sanitize_paths?, omit_raw_paths?) -> {missing, missing_count, diagnosis} build_relink_plan(search_roots, max_depth?, max_seconds?, max_files_scanned?, skip_search_when_volume_missing?, sanitized?) -> {candidates, diagnosis} conform_boundary_report(...) -> {capabilities, timeline, gaps_overlaps, source_ranges, missing_media} audio_capabilities() -> {supported, partially_supported, unsupported} probe_audio_item(track_type?, track_index?, item_index?) -> {summary, audio_properties, source_audio_mapping} probe_audio_track(track_index?) -> {track_count, enabled, locked, sub_type, voice_isolation} safe_set_audio_properties(properties, restore?, dry_run?, track_type?, track_index?, item_index?) -> {success, results} audio_mix_capability_report(...) -> {capabilities, mix_recommendations} voice_isolation_capabilities(track_index?, track_type?, item_index?) -> {timeline_track, item} audio_mapping_report(clip_ids?) -> {timeline_items, media_pool_items} safe_auto_sync_audio(clip_ids|selected, settings?, dry_run?) -> {success} transcription_capabilities(clip_ids?|selected?) -> {clip_methods, folder} subtitle_generation_probe(settings?, allow_generate?) -> {success} fairlight_boundary_report(...) -> {capabilities, track, item, audio_mapping, transcription} For long-form per-action guidance and a worked example, call: timeline(action="action_help", params={"name": "<action>"})

Reveal a file path in Resolve's Media Storage browser. Args: file_path: Absolute path to the file to reveal.

Import/export render and burn-in presets. Actions: import_render(path) -> {success} export_render(name, path) -> {success} import_burnin(path) -> {success} export_burnin(name, path) -> {success} list_burnin() -> {presets} — Resolve 21.0.4+; burn-in preset names usable with the render tool's DataBurnIn setting and load_burnin_preset on project_settings / timeline_item delete_burnin(name) -> {success} — Resolve 21.0.4+

Manage DaVinci Resolve projects. Actions: list() -> {projects} list_attributes() -> {projects: {name: {lastModifiedDate, creationDate, notes, liveCollaborationMode}}} — Resolve 21.0.4+. Per-project attributes for the current folder without loading any project. get_current() -> {name, id} create(name, media_location_path?) -> {success, name} load(name) -> {success} save() -> {success} close() -> {success} delete(name) -> {success} import_project(path, name?) -> {success} export_project(name, path, with_stills_and_luts?) -> {success} archive(name, path, src_media?, render_cache?, proxy_media?) -> {success} restore(path, name?) -> {success} project_capabilities() -> {capabilities} probe_project_lifecycle() -> {project_manager, ...} probe_project_settings(keys?, try_write?) -> {snapshot, candidate_settings} safe_project_create(name, media_location_path?, dry_run?) -> {success} safe_project_export(name, path, with_stills_and_luts?, dry_run?) -> {success} safe_project_import(path, name, dry_run?) -> {success} safe_project_archive(name, path, src_media=false, render_cache=false, proxy_media=false, dry_run?) -> {success} safe_project_restore(path, name, dry_run?) -> {success} safe_project_delete(name, close_current?, dry_run?) -> {success} safe_set_project_settings(settings, restore?, dry_run?) -> {success} project_settings_snapshot(name?) -> {project, settings, presets, ...} database_capabilities() -> {methods, current, databases} safe_set_current_database(db_info, dry_run?, allow_switch?) -> {success} preset_lifecycle_probe() -> {project_presets, render_presets, layout_presets, ...} project_boundary_report() -> {capabilities, project_manager, settings, database, presets, cloud} lint() -> {ok, counts, issues} — graded project health pre-flight (no project, no current timeline, mixed fps, empty timeline, render/color-science unset, offline media). diff_to_spec(spec_path|spec) -> {actions, diff, change_count} — preview drift vs a declarative spec WITHOUT mutating. Spec is YAML/JSON: {project, color_preset?, settings?, timelines:[{name,fps?,settings?,markers?}], hooks?}. plan_spec(spec_path|spec) -> {dry_run, actions, diff, change_count} — same as apply with dry_run. apply_spec(spec_path|spec, dry_run?, run_hooks?, continue_on_error?) -> {success, applied, failures} Reconcile the project toward the spec (idempotent: re-runs are no-ops). Color/HDR settings are applied in dependency order; markers only added if absent. Hooks run only when run_hooks=true (executes shell from the spec — opt-in).

Navigate and manage project folders in the Project Manager. Actions: list() -> {folders} get_current() -> {folder} create(name) -> {success} delete(name) -> {success} open(name) -> {success} goto_root() -> {success} goto_parent() -> {success}

Cloud project operations (requires DaVinci Resolve cloud infrastructure). Actions: create(settings) -> {success} — settings: {CLOUD_SETTING_PROJECT_NAME, ...} load(settings) -> {success} import_project(path, settings) -> {success} restore(folder_path, settings) -> {success}

Manage DaVinci Resolve project databases. Actions: get_current() -> {db_type, db_name} list() -> {databases} set_current(db_info) -> {success} — db_info: {DbType, DbName}

Project metadata, settings, and color groups. Actions: get_name() -> {name} set_name(name) -> {success} get_setting(name?) -> {settings} — omit name for all settings set_setting(name, value) -> {success, known_limitation?} A refusal carries the api_truth entry for that key when one exists — several settings cannot be written from the API at all. get_unique_id() -> {id} get_presets() -> {presets} set_preset(name) -> {success} refresh_luts() -> {success} get_gallery() -> {available} export_frame_as_still(path) -> {success} project_summary(include_clips?, clip_limit?) -> {project, current_page, timeline_count, current_timeline, media_pool} Live structural readout — what's in this project right now (no analysis needed). load_burnin_preset(name) -> {success} insert_audio(media_path, start_offset?, duration?) -> {success} get_color_groups() -> {groups} add_color_group(name) -> {success, name} delete_color_group(name) -> {success} apply_fairlight_preset(preset_name) -> {success} generate_speech(speech_generation_settings, timecode?) -> {success, new, new_id} — Resolve 21+, AI Speech Generator; creates new audio media (confirm-gated) reset_intellisearch_analysis() -> {success} — Resolve 21+; clears the project's IntelliSearch analysis data

Render pipeline: jobs, presets, formats, codecs, and rendering. Actions: add_job() -> {job_id} delete_job(job_id) -> {success} delete_all_jobs() -> {success} list_jobs() -> {jobs} get_job_status(job_id) -> {status} start(job_ids?, interactive?) -> {success} stop() -> {success} is_rendering() -> {rendering} get_formats() -> {formats} get_codecs(format) -> {codecs} get_format_and_codec() -> {format, codec} set_format_and_codec(format, codec) -> {success} get_mode() -> {mode} set_mode(mode) -> {success} get_resolutions(format, codec) -> {resolutions} get_settings() -> {settings} (alias for set_render_settings with get) set_settings(settings) -> {success, ignored_settings?, warnings?} Resolve 21.0.4+ settings keys: UseFullExtents (bool), AddFrameHandles (int >= 0), DataBurnIn (burn-in preset name, "Same as project", or "None"). AddFrameHandles is ignored while UseFullExtents is true — the call still succeeds, so that pairing comes back in warnings rather than silently doing nothing. list_presets() -> {presets} load_preset(name) -> {success} save_preset(name) -> {success} delete_preset(name) -> {success} quick_export_presets() -> {presets} quick_export(preset, params?) -> {status} render_capabilities() -> {methods, formats, presets, quick_export_presets} probe_render_matrix(formats?, max_pairs?) -> {matrix, errors} probe_render_settings() -> {format_and_codec, mode, settings, jobs, is_rendering} validate_render_settings(settings, require_temp_target?) -> {valid, errors, unknown_keys} safe_set_render_settings(settings, dry_run?, restore?, require_temp_target?) -> {success, diff} prepare_render_job(target_dir, settings?, format?, codec?, custom_name?, from_preset?, dry_run?) -> {success, job_id} from_preset pins the base render state (LoadRenderPreset) before the explicit settings go on top. Without it the job inherits the Deliver page's loaded preset for every key not passed, which the API gives no way to read back — see the SetRenderSettings api_truth entry. render_job_lifecycle_probe(target_dir, settings?, format?, codec?, custom_name?) -> {success, job_id, status_before_delete} quick_export_capabilities() -> {presets, safe_params, guards} safe_quick_export(preset, target_dir?|params?, custom_name?, dry_run?, allow_render?) -> {success, status} export_render_boundary_report(include_matrix?, max_pairs?, include_quick_export?) -> {capabilities, settings, matrix?} list_delivery_targets(tier?, check_availability?) -> {targets, tiers, schema_version} resolve_delivery_target(target, overrides?) -> {format_id, codec_id, settings, qc_spec} prepare_delivery_job(target, target_dir, overrides?, settings?, custom_name?, dry_run?) -> {success, job_id, qc_spec} Delivery targets are named render intents (`prores422hq_master`, `youtube`, `tiktok`, ...). One definition emits both the Resolve render settings and an ffprobe-shaped QC spec, so the advanced server's `deliverable_qc` can verify a rendered file against the same intent that produced it. Format/codec are resolved against the live matrix, so an intent unavailable on this machine/license fails with the actual available lists.

Get subfolders in a given absolute folder path from Media Storage. Args: folder_path: Absolute path to the folder to list subfolders for.

Rename the current project. Args: name: New name for the project.

Browse storage volumes and import media into the Media Pool. Actions: get_volumes() -> {volumes} get_subfolders(path) -> {subfolders} get_files(path) -> {files} reveal(path) -> {success} import_to_pool(items) -> {imported} — simple: params.items is a list of absolute file/folder paths import_to_pool(item_infos) -> {imported} — positioned: params.item_infos is a list of {media, startFrame, endFrame} dicts per docs line 210. Mirrors MediaStorage.AddItemListToMediaPool([{itemInfo}, ...]). add_clip_mattes(clip_id, paths, stereo_eye?) -> {success} add_timeline_mattes(paths) -> {items}

Manage the Media Pool: folders, clips, timelines, import/export. PREFER safe_* / probe_* / *_capabilities / *_boundary_report / *_checked variants where they exist. Raw mutators below do not validate paths, support dry_run, or normalize errors. Actions: get_root_folder() -> {name, id} get_current_folder() -> {name, id} set_current_folder(path) -> {success} — path like "Master/SubFolder" add_subfolder(name, parent_path?) -> {success, name, id} delete_folders(folder_ids) -> {success} DESTRUCTIVE. Deletes folders + every clip they contain. move_folders(folder_ids, target_path) -> {success} refresh() -> {success} create_timeline(name, if_exists?) -> {success, name, id} — if_exists: version (default), reuse, or fail create_timeline_from_clips(name, clip_ids, if_exists?) -> {success, name, id} — simple: params.clip_ids appends clips end-to-end into a new timeline create_timeline_from_clips(name, clip_infos, if_exists?) -> {success, name, id} — positioned: params.clip_infos is a list of {clip_id or media_pool_item_id, start_frame & end_frame (or startFrame/endFrame), record_frame/recordFrame}. record_frame is relative to the created timeline start frame by default; pass record_frame_mode="absolute" for raw Resolve recordFrame values. setup_multicam_timeline(name, clip_ids|angles, sync_mode?, include_audio?, dry_run?) -> {success} — creates a stacked multicam prep timeline: one angle per video track, optional matching audio tracks. Native multicam clip conversion remains a Resolve UI step. import_timeline(path, options?, background?) -> {success, name | job_id} UNSAFE. No path sandboxing. Prefer timeline.import_timeline_checked. delete_timelines(timeline_ids) -> {success} CATASTROPHIC. Deletes timelines outright. append_to_timeline(clip_ids) -> {success, count} — legacy: params.clip_ids only (appends at end / default placement) append_to_timeline(clip_infos) -> {success, count, items} — positioned: params.clip_infos is a list of {clip_id or media_pool_item_id, start_frame & end_frame (or startFrame/endFrame), record_frame/recordFrame, track_index/trackIndex (1-based), optional media_type/mediaType (1=video, 2=audio)}. record_frame is relative to the current timeline start frame by default; pass record_frame_mode="absolute" for raw Resolve recordFrame values. Returns timeline_item_id per item. import_media(paths) -> {imported} UNSAFE. No dry_run. Prefer safe_import_media. — simple: params.paths is a list of file/folder paths import_media(clip_infos) -> {imported} UNSAFE. No dry_run. Prefer safe_import_sequence. — image sequences: params.clip_infos is a list of {FilePath, StartIndex, EndIndex} dicts (PascalCase keys per Resolve docs). Example: [{"FilePath": "frame_%03d.dpx", "StartIndex": 1, "EndIndex": 100}] delete_clips(clip_ids) -> {success} DESTRUCTIVE. Removes clips from the Media Pool (does not touch source files). move_clips(clip_ids, target_path) -> {success} relink(clip_ids, folder_path) -> {success} UNSAFE. No dry_run. Prefer safe_relink. unlink(clip_ids) -> {success} UNSAFE. No dry_run. Prefer safe_unlink. export_metadata(path, clip_ids?) -> {success} get_unique_id() -> {id} create_stereo_clip(left_id, right_id) -> {success, name} auto_sync_audio(clip_ids, settings?) -> {success} get_selected() -> {clips} set_selected(clip_id) -> {success} get_clip_mattes(clip_id) -> {mattes} get_timeline_mattes(folder_path?) -> {mattes} delete_clip_mattes(clip_id, paths) -> {success} import_folder(path, source_clips_path?) -> {success} ingest_capabilities() -> {supported, partially_supported, unsupported} probe_media_pool(depth?) -> {media_pool_id, methods, root, current_folder, selected_clips} depth defaults to 1, max 4. Folders below the cutoff have truncated:true and a real subfolder_count; re-probe deeper or use folder.get_subfolders. probe_ingest_item(clip_ids? selected?) -> {items, count} safe_import_media(paths, target_folder?, dry_run?) -> {success, imported, clips} safe_import_sequence(FilePath|file_path|pattern, StartIndex?, EndIndex?, target_folder?, dry_run?) -> {success, imported, clips} safe_import_folder(path, source_clips_path?, dry_run?) -> {success} organize_clips(clip_ids|selected, target_path, create_missing?, dry_run?) -> {success} copy_metadata(source_clip_id, target_clip_ids, keys?, include_third_party?, dry_run?) -> {success, results} normalize_metadata(clip_ids|selected, metadata?, third_party_metadata?, dry_run?) -> {success, results} probe_clip_properties(clip_ids|selected) -> {items, count} metadata_field_inventory(clip_ids|selected, include_values?) -> {items, ui_group_names} safe_relink(clip_ids|selected, folder_path, dry_run?) -> {success} safe_unlink(clip_ids|selected, dry_run?) -> {success} check_proxy_media_compatibility(clip_id, proxy_path|path, expected_codec?, expected_profile?) -> {compatible, mismatches} link_proxy_checked(clip_id, proxy_path|path, dry_run?, check_compatibility?, require_compatible?, expected_codec?, expected_profile?) -> {success} link_full_resolution_checked(clip_id, path|full_res_media_path, dry_run?) -> {success} set_clip_marks(clip_ids|selected, mark_in, mark_out, type?, dry_run?) -> {success, results} clear_clip_marks(clip_ids|selected, type?, dry_run?) -> {success, results} copy_clip_annotations(source_clip_id, target_clip_ids, include_markers?, include_flags?, include_clip_color?, dry_run?) -> {success, results} media_pool_boundary_report(depth?, clip_ids?, selected?) -> {capabilities, media_pool, items?}

Operations on Media Pool folders. Address a folder with `path` ("Master/SubFolder") or with `folder_id` (the id get_subfolders returns). Omit both for the current folder. An address that is supplied but does not resolve is an error — it never falls back to the current folder. Actions: get_clips(path?) -> {clips} — path like "Master/SubFolder", omit for current get_name(path?) -> {name} get_subfolders(path?) -> {subfolders} is_stale(path?) -> {stale} get_unique_id(path?) -> {id} export(path?, export_path) -> {success} transcribe_audio(path?, use_speaker_detection?, background?) -> {success | job_id} — use_speaker_detection is Resolve 21+; background=true returns a job_id (poll resolve_control job_status) clear_transcription(path?) -> {success} perform_audio_classification(path?) -> {success} — Resolve 21+ clear_audio_classification(path?) -> {success} — Resolve 21+ analyze_for_intellisearch(path?, identify_faces?, is_better_mode?) -> {success} — Resolve 21+, AI IntelliSearch Extra analyze_for_slate(path?, marker_color?) -> {success} — Resolve 21+, AI Slate ID Extra remove_motion_blur(path?, deblur_option?) -> {success, created} — Resolve 21+; renders NEW media (confirm-gated)

Operations on a media pool clip. Identify clip by clip_id. Actions: get_name(clip_id) -> {name} get_metadata(clip_id, key?) -> {metadata} — key: one string, or a list of strings to get just that subset. Missing keys: the list form maps them to null (distinguishing absent from empty); the string form passes Resolve's own answer through unchanged, which is "" or null depending on the getter and build (get_clip_property returns null on Studio 19.1.3.7). set_metadata(clip_id, key, value) OR set_metadata(clip_id, metadata) -> {success} get_third_party_metadata(clip_id, key?) -> {metadata} — key: one string, or a list of strings to get just that subset. Missing keys: the list form maps them to null (distinguishing absent from empty); the string form passes Resolve's own answer through unchanged, which is "" or null depending on the getter and build (get_clip_property returns null on Studio 19.1.3.7). set_third_party_metadata(clip_id, key, value) -> {success} get_media_id(clip_id) -> {media_id} get_clip_property(clip_id, key?) -> {properties} — key: one string, or a list of strings to get just that subset. Missing keys: the list form maps them to null (distinguishing absent from empty); the string form passes Resolve's own answer through unchanged, which is "" or null depending on the getter and build (get_clip_property returns null on Studio 19.1.3.7). set_clip_property(clip_id, key, value) -> {success} get_clip_color(clip_id) -> {color} set_clip_color(clip_id, color) -> {success} clear_clip_color(clip_id) -> {success} link_proxy(clip_id, proxy_path) -> {success} unlink_proxy(clip_id) -> {success} replace_clip(clip_id, path) -> {success} set_name(clip_id, name) -> {success} link_full_resolution_media(clip_id, path) -> {success} monitor_growing_file(clip_id) -> {success} replace_clip_preserve_sub_clip(clip_id, path) -> {success} get_unique_id(clip_id) -> {id} transcribe_audio(clip_id, use_speaker_detection?, background?) -> {success | job_id} — use_speaker_detection is Resolve 21+; background=true returns a job_id (poll resolve_control job_status) clear_transcription(clip_id) -> {success} get_transcription(clip_id) -> {text, truncated, status, has_transcription} Read a clip's transcription. `truncated` flags when Resolve's preview property cut the text off (the full transcript is longer). Clip-level and separate from the timeline subtitle transcript (timeline.get_transcript) that propose_cuts uses. extract_frames(clip_id, timestamps, output_dir?) -> {frame_paths, output_dir, count, errors} Extract still JPEGs from the clip's source at the given timestamps (seconds) via ffmpeg. Source-safe: reads source, writes only to a scratch dir. perform_audio_classification(clip_id) -> {success} — Resolve 21+ clear_audio_classification(clip_id) -> {success} — Resolve 21+ analyze_for_intellisearch(clip_id, identify_faces?, is_better_mode?) -> {success} — Resolve 21+, AI IntelliSearch Extra analyze_for_slate(clip_id, marker_color?) -> {success} — Resolve 21+, AI Slate ID Extra remove_motion_blur(clip_id, deblur_option?) -> {success, new, new_id} — Resolve 21+; renders NEW media (confirm-gated) get_audio_mapping(clip_id) -> {mapping} get_mark_in_out(clip_id) -> {mark} set_mark_in_out(clip_id, mark_in, mark_out, type?) -> {success} clear_mark_in_out(clip_id, type?) -> {success} get_timeline(clip_id) -> {is_timeline, timeline: {name, unique_id?, start_frame?, end_frame?}} — Resolve 21.0.4+. Resolves a Media Pool timeline entry to its timeline object summary; is_timeline=false for ordinary clips. open_in_viewer(clip_id, page?, mark_in_seconds?, mark_out_seconds?, clear_marks?) -> {success, clip_id, clip_name, folder_name, page, mark_set} — Switches to Media page (default) and selects the clip in the bin. Resolve auto-loads the selected clip into the source viewer on Media page. Optional: pre-set Mark In/Out on the clip so a shot's time range is highlighted when it loads. NOTE: Source-viewer auto-load is undocumented in the Scripting API; verified empirically on Resolve Studio 20.3.2.9. If BMD changes the behavior the clip will still be selected — editor double-clicks to load.

Markers and flags on media pool clips. Identify clip by clip_id. Actions: add(clip_id, frame|frame_id|frameId, color?, name?, note?, duration?, custom_data?) -> {success, frame} get_all(clip_id) -> {markers} get_by_custom_data(clip_id, custom_data) -> {markers} update_custom_data(clip_id, frame|frame_id|frameId, custom_data) -> {success} get_custom_data(clip_id, frame|frame_id|frameId) -> {data} delete_by_color(clip_id, color) -> {success} delete_at_frame(clip_id, frame|frame_id|frameId) -> {success} delete_by_custom_data(clip_id, custom_data) -> {success} add_flag(clip_id, color) -> {success} get_flags(clip_id) -> {flags} clear_flags(clip_id, color) -> {success} set_name(clip_id, name) -> {success} link_full_resolution_media(clip_id, path) -> {success} monitor_growing_file(clip_id) -> {success} replace_clip_preserve_sub_clip(clip_id, path) -> {success}

Project-scoped media analysis and guarded metadata publishing. <when_to_use> - Lead any analysis flow with capabilities() + get_caps() to know what's available and what the current budget looks like. - For per-clip work: analyze_clip(clip_id=...). For bins/projects/timelines: analyze_bin, analyze_project, analyze_sequence (return job_id for long runs). - Vision uses host_chat_paths: analyze returns frame_paths + JSON schema; the host reads frames, produces JSON, then calls commit_vision to merge. - For caps refusal recovery: set_caps_preset (preset=generous|unlimited) or wait for day rollover. </when_to_use> <returns> capabilities -> {tools, transcription, vision} get_caps -> {preset, caps, presets_available, usage?} set_caps_preset -> {success, preset, overrides} plan -> {clips, artifacts, estimated_seconds} analyze_clip|file -> {clips, manifest: {success, clip_count, successful_clip_count, failed_clip_count, partial_success?, caps_refusal_clip_count?, clips: [...], error?}} analyze_bin|project|sequence -> {job_id, status, clips, manifest, ...} commit_vision -> {analysis_json, marker_plan_json, metadata_publish} summarize -> {clips_summarized, summary, provenance: {source_reports, missing_reports}} review_timeline_markers -> {path, samples, vision_review?} start_batch_job -> {job_id, status} batch_job_status -> {job_id, status, progress, recent_events, clip_states} All actions may return {"error": {code, category, retryable, message, remediation, reason?}}. Errors with code=CAPS_REFUSAL also carry a `caps_refusal` block at the clip level and a `caps_refusal_clip_count` aggregate at the manifest level. </returns> Actions: capabilities() -> {tools, transcription, vision} install_guidance() -> {missing} — guidance only; never installs anything get_caps(clip_id?, job_id?) -> {preset, caps, presets_available, usage?} — effective caps + usage rollup (vision tokens consumed per scope, percent of budget). set_caps_preset(preset, overrides?) -> {success, preset, overrides} — preset is minimal | standard | generous | unlimited. Overrides is a dict of {field: int|"unlimited"} that wins over the preset. get_usage(scope?, scope_key?, clip_id?, job_id?) -> {scope, usage} — raw usage rollup for one scope (clip | job | day). get_resolve_ai_usage(session_only?, op?, limit?) -> {summary, recent} — ledger of Resolve 21 local AI ops (audio classification, IntelliSearch, slate, motion-deblur, speech). Tracks invocations, wall-clock, and files/bytes created by the two media-creating ops. Separate from get_caps/get_usage (those meter Claude-side tokens; these ops don't spend them). get_ai_governance() -> {tier, mode, thresholds, usage, tiers_available} — governance tier (off|lenient|standard|strict) for the media-creating AI ops vs this session's render usage. mode is advisory (default; preview warnings only) or enforce (over-tier runs are blocked). set_ai_governance(preset?, mode?, overrides?) -> {success, tier, mode, overrides} — set the tier, the mode (advisory|enforce), and/or overrides (deblur_runs, speech_runs, render_bytes, render_wall_clock_ms; int or "unlimited"). In enforce mode a blocked run returns GOVERNANCE_BLOCKED; pass override_governance=true on the op to consciously exceed the tier once. resolve_output_root(analysis_root?, source_paths?) -> {project_root} plan(target, depth?, analysis_root?, transcription?, vision?, dry_run?) -> {clips, artifacts} analyze_file(path|file_path, dry_run?, session_only?, persist?, prefer_handle?|background?|async_job?) -> {clips, manifest} | {success, job, plan, running, note} when async analyze_clip(clip_id|selected, dry_run?, session_only?, persist?, prefer_handle?|background?|async_job?) -> {clips, manifest} | {success, job, plan, running, note} when async analyze_bin(path|bin_path, recursive?, dry_run?, session_only?, persist?, prefer_handle?|background?|async_job?) -> {clips, manifest} | {success, job, plan, running, note} when async analyze_project(recursive?, dry_run?, session_only?, persist?, prefer_handle?|background?|async_job?) -> {clips, manifest} | {success, job, plan, running, note} when async analyze_sequence(timeline_index?, track_types?, dry_run?, session_only?, persist?, prefer_handle?|background?|async_job?) -> {clips, manifest} | {success, job, plan, running, note} when async analyze_timeline(...) -> alias for analyze_sequence on the current timeline -- async opt-ins on the analyze_* actions above. Both reroute to start_batch_job and return its {success, job, plan} envelope — the id is job.job_id, NOT a top-level job_id — plus `running` and a `note` naming the next call. They differ in what happens next: prefer_handle=true job is created and left queued; nothing runs until you call run_batch_job_slice yourself. Unchanged contract. background|async_job=true job is created AND driven to completion off-thread, matching what `background` means on every other tool here. Poll batch_job_status until status is completed / completed_with_errors / canceled. An explicit dry_run=true still returns the synchronous plan and starts nothing. A dry_run that came from the dry_run_first_default preference does not override an explicit async request. detect_sync_events(paths?|target?, event_types?, windows?) -> {files, alignment} add_sync_event_markers(target?|paths?|detections?, confirm?) -> {added, skipped} publish_clip_metadata(target?, fields?, slate_detection?, timed_markers?|write_markers?, dry_run?, confirm?) -> {results} commit_vision(clip_id|file_path|clip_dir, visual, vision_token?, analysis_root?, publish_metadata?, dry_run?, confirm?) -> {analysis_json, marker_plan_json, metadata_publish} review_timeline_markers(max_samples?, analysis_root?, vision?) -> {path, samples, vision_review?} summarize(analysis_root?) -> project summary from existing clip reports get_report(report_path?) -> load manifest or a report under the analysis root build_index(analysis_root?) -> rebuild the single-user SQLite index from existing reports index_status(analysis_root?) -> summarize index path, size, and row counts query_index(query, limit?, result_types?) -> search clips, markers, and transcripts start_batch_job(target, depth?, ...) -> create a durable slice-runnable analysis job; target may be selected, bin, project, sequence, file, or {type:"clips", clip_ids:[...]} run_batch_job_slice(job_id, max_clips?, max_seconds?) -> process a bounded job slice and refresh the index after successful clips batch_job_status(job_id) -> progress, recent events, and clip states list_batch_jobs() -> recent durable analysis jobs cancel_batch_job(job_id) -> stop future slices resume_batch_job(job_id) -> requeue a canceled or interrupted job cleanup_artifacts(frames_only=true) -> remove generated frame artifacts only strata_status(clip_id?) -> perception-strata inventory (events/curves/words/story beats) for the project or one clip backfill_words(analysis_root?) -> promote word timestamps from stored report blobs into transcript_words rows strata_run(clip_id, analyzers?) -> run local strata analyzers (prosody | beat_grid | motion_energy | face) on one clip; default = whatever this machine can run take_diff(clip_a, clip_b, text?) -> align two takes on transcript words and diff their delivery (pace, pauses, pitch, energy) — measures only, never picks cut_candidates(clip_id, time_seconds, window_seconds?, fps?, limit?) -> rank cut frames around an intended joint with reasons (blink/pause/word-gap/breath/beat/motion) — aims, never decides strata_query(clip_id?, start_seconds?, end_seconds?, match_word?, context_seconds?, include_curve_values?, limit?) -> cross-track window bundle for a clip, or project-wide word find with joined context per hit timeline_strata(timeline_name, timeline_version?, record_start_frame?, record_end_frame?, fps?) -> project clip strata through a timeline's recorded placements (whole-clip bundles; frames are absolute record frames incl. start-timecode offset; fps defaults to the snapshot's stored value; source-offset mapping needs the live item) plan_story_beats(clip_id) -> timecoded transcript digest + prosody evidence + JSON schema for the host model to segment into story beats (no LLM call server-side) commit_story_beats(clip_id, beats, source_model?) -> validate + persist host-produced story beats; machine commits supersede machine rows only, human rows always win list_story_beats(clip_id) -> current (non-superseded) story beats for a clip Analysis outputs stay under a davinci-resolve-mcp-analysis project root and are validated so they are never written beside source media. Executed Resolve-target analysis defaults to execution, persisted inspectable artifacts, metadata writeback, and Media Pool clip markers unless disabled. The SQLite index is a single-user derived cache stored beside the JSON reports; it stores text/metadata only, never sampled image bytes. Batch jobs are single-user operational state stored beside the analysis root; each slice exits cleanly so agents can poll progress without spending tokens on long media processing. Executed Resolve-target analysis writes metadata and Media Pool clip markers by default; pass dry_run=true, publish_metadata=false, or timed_markers=no to disable writeback. If ask_before_metadata_publish is enabled, publish_clip_metadata returns a confirmation prompt until confirm=true. Vision uses host_chat_paths by default: analyze actions return a deferred payload with absolute frame_paths and a JSON schema. The host chat reads each frame as a local image, produces JSON per the schema, and calls media_analysis(action="commit_vision", params={clip_id, visual, vision_token}) to merge the result, rebuild Media Pool clip markers, and publish vision-dependent metadata to Resolve. Works with any MCP client whose chat model is vision-capable; no sampling/createMessage support required.

Timeline version-on-mutate, archive, rollback, and brain-edit history (C6). Every destructive timeline op auto-archives the working timeline to the Archive bin under an analysis_run_id; these actions let you inspect and control that. Actions: begin_run(label?, initiator?, analysis_run_id?) -> {success, analysis_run_id, label, initiator, started_at} Open a run. Subsequent destructive calls without an explicit analysis_run_id auto-thread this one — so a multi-step brain operation creates ONE archive instead of N. Pair with end_run to capture the cumulative metric delta. end_run(analysis_run_id?) -> {success, analysis_run_id, ended_at, summary} Close the active run (or a specific one). Aggregates brain_edits into per-metric rollup in analysis_runs.summary_json. list_runs(limit?) -> {success, runs} Recent runs with their summaries, newest first. archive_current(reason?, analysis_run_id?) -> {success, timeline_name, archived_timeline_name, version, archive_bin, row_id} Manually checkpoint the current timeline. If analysis_run_id is supplied and that run already archived the current timeline, this is a no-op. list_versions(timeline_name) -> [{version, archived_timeline_name, created_at, ...}] Version chain for `timeline_name`, oldest first. Includes any retention- collapsed versions (drt_export_path populated). diff_timelines(from_timeline, to_timeline) -> {added, removed, moved, trimmed, summary} Structural diff between two LIVE timelines by name (read-only; no archived versions needed). Built for edit-engine variants — tighten/ selects produce new-name timelines with no shared version chain. get_history(timeline_name?, analysis_run_id?, limit?) -> [{edit_type, target_metric, before_value, after_value, delta, ...}] Brain-edit history. Filter by timeline_name or analysis_run_id; defaults to the most recent 50 across the project. rollback(timeline_name, version, analysis_run_id?) -> {success, restored_timeline_name, archive_of_previous} Restore an archived version. Archives current state first; restored copy gets a "_rolled_back_<HHMMSS>" suffix. prune(timeline_name, keep_n=10) -> {success, pruned, kept, details} Collapse old versions to .drt exports under _soul/timeline_versions/<slug>/, delete the archived timeline from the bin, keep the DB row for rollback. registry() -> {entries, registry_path} Read the cross-project brain_edits registry that lives one level above each project root (analogous to analysis_registry.json).

Evidence-driven edit loops (Phase E): selects assembly, tighten, swap. Every loop is plan → confirm → execute. plan_* actions are dry-run by construction: they query the DB-canonical analysis store and return a per-decision rationale plus a stored plan_id. execute_* actions require a confirm_token, run under the version-on-mutate hook (the timeline is archived first), and return before/after metrics as readback. Actions: - plan_selects(min_select_potential?, max_duration_seconds?, timeline_name?, max_shots?, handle_seconds?, analysis_root?) — rank shots by deep-tier select potential (clip-level fallback), story-spine order. Additive. - execute_selects(plan_id, confirm_token?) — creates a NEW selects timeline from the plan's clip in/out ranges. Nothing existing is touched. - plan_tighten(timeline_name?, target_ratio?, min_pause_seconds?, handle_seconds?, include_audio?) — dead-air lifts from transcript-gap evidence for the current (or named) timeline. Kept ranges mirror onto the items' linked audio tracks by default so the variant is audible; pass include_audio=false for a video-only (silent) assembly. - execute_tighten(plan_id, confirm_token?, include_details?) — assembles a tightened VARIANT timeline from the plan's keep ranges (video + mirrored audio), never mutating the original. Readback includes an audio_accounting block and a compact structural_diff (counts + a small sample); the full per-item diff is persisted in the plan record (get_plan → execution_summary) and returned inline only when include_details=true. - plan_silence_ripple(timeline_name?, track_index?, threshold_db?, min_strip_frames?, pre_head_frames?, post_tail_frames?, include_audio?) — waveform silence strips via ffmpeg silencedetect (Resolve's Ripple Delete Silence parity). Items without readable file paths ride along whole (reported in skipped). - execute_silence_ripple(plan_id, confirm_token?, include_details?) — same variant assembly as execute_tighten; original timeline untouched. - plan_swap(track_index?, timeline_start_frame | item_name, kind?, limit?) — alternates for one timeline item via the similarity index. - execute_swap(plan_id, alternate_index, confirm_token?) — replaces the item in place (lift + positioned append, same slot) on the current timeline (version-archived first). - list_plans(limit?) / get_plan(plan_id)

Markers and playhead operations on the current timeline. Marker frames are RELATIVE to the timeline start: frame 0 is the first frame of the timeline, even when the timeline starts at 01:00:00:00. Marker timecode params are absolute timeline timecode as shown in the Resolve UI (timecodes before the start timecode are treated as elapsed time) and are converted to relative frames automatically. set_current_timecode accepts the same convention: elapsed timecodes below the start timecode are lifted to absolute before calling Resolve, which itself refuses sub-start timecodes with a bare False. Actions: add(frame|frame_id|frameId|timecode?, color?, name?, note?, duration?, custom_data?) -> {success, frame} If frame/timecode is omitted, add uses the current playhead timecode. get_all() -> {markers} get_by_custom_data(custom_data) -> {markers} update_custom_data(frame|frame_id|frameId|timecode, custom_data) -> {success} get_custom_data(frame|frame_id|frameId|timecode) -> {data} delete_by_color(color) -> {success} delete_at_frame(frame|frame_id|frameId|timecode) -> {success} delete_by_custom_data(custom_data) -> {success} get_current_timecode() -> {timecode} set_current_timecode(timecode) -> {success} get_current_video_item() -> {name, id} get_thumbnail() -> {thumbnail} get_thumbnail_image() -> MCP image content for the current Color-page frame annotation_capabilities() -> {scopes, marker_colors, frame_aliases} probe_annotations(scope?, ...) -> {scopes, count} normalize_marker_payload(frame|timecode?, color?, name?, note?, duration?, custom_data?) -> {marker} copy_annotations(source={scope,...}, target={scope,...}, include_flags?, include_clip_color?) -> {success} move_annotations(source={scope,...}, target={scope,...}) -> {success} sync_marker_custom_data(scope?, frame|timecode, custom_data, ...) -> {success} clear_annotations_by_scope(scope?, color?, custom_data?, all?, clear_flags?, clear_clip_color?) -> {success} export_review_report(scope?, include_capabilities?) -> {title, annotations, capabilities?} annotation_boundary_report(scope?) -> {capabilities, annotations}

See what Resolve is rendering — capture a timeline frame as a viewable image. <when_to_use> - Verifying anything visual: title placement and safe area, framing, a grade, a Fusion comp, a transition, an artefact. Read the frame instead of inferring from metadata. - Confirming an edit landed where you meant it — capture at the cut timecode. - Before and after a change, at the same timecode, to show what moved. </when_to_use> Captures Resolve's processed output — grade, Fusion, titles, transitions — not the source file. For the raw camera file use media_analysis(action="extract_frames"). Actions: capture(timecode?|frame?, quality?, max_width?, format?, timeline_name?) -> MCP image content capabilities() -> {quality_modes, ffmpeg, render_settings_restorable, ...} capture parameters: timecode Absolute ('01:00:15:12') or elapsed ('00:00:15:12') timeline timecode. Omit to capture the current playhead. frame Alternative to timecode: absolute timeline frame number. quality 'frame' (default) renders exactly that frame — the only frame-accurate route, full resolution, ~1s, works headless. 'preview' is the same render bounded to max_width 1280. 'thumbnail' is instant and touches nothing, but returns the CLIP's thumbnail (see below). 'still' uses a Gallery still. max_width Cap the width in pixels to conserve context. Needs ffmpeg on the render path; without it the call fails rather than quietly returning a full-size frame. format jpg (default), png, or tif. timeline_name Capture from a different timeline; it is made current for the read and the original is restored afterwards. Choosing a quality — the trade-off is accuracy against side effects: 'frame'/'preview' Frame-exact. Renders one frame, so it changes project-level render settings. Format and codec are restored; TargetDir, CustomName and the mark range cannot be read back on builds without GetRenderSettings, so they are reset to the full timeline rather than truly restored. Refuses while another render is running. 'thumbnail' Changes nothing and returns instantly, but it is NOT frame accurate: GetCurrentClipThumbnailImage returns the clip's thumbnail, identical for every frame of that clip (measured on Studio 19.1.3.7). Use it to see which clip is under the playhead, never to judge a specific frame. It also needs the Color page AND Resolve frontmost, or it returns nothing. 'still' Full-resolution Gallery still. Requires the Gallery panel to be open on the Color page — no scripting call can open it, so this fails with a bare refusal when it is closed. The playhead, the Color page, the current timeline and the Gallery are all restored; a capture is a read of the picture, not an edit of the cut.

AI and analysis operations on the current timeline. Actions: create_subtitles(settings?, background?) -> {success | job_id} — auto-caption from audio. These are long ops: background=true returns a job_id immediately; poll resolve_control(action="job_status", params={"job_id": ...}). detect_scene_cuts(background?) -> {success | job_id} analyze_dolby_vision(clip_ids?, analysis_type?, background?) -> {success | job_id} grab_still() -> {success} grab_all_stills(source?) -> {count}

Properties, transforms, speed, keyframes, and metadata for a timeline item. Identify by track_type, track_index, item_index (item_index is 0-BASED: 0 = first clip; track_index is 1-based). Actions: get_name(track_type?, track_index?, item_index?) -> {name} get_property(key?, ...) -> {properties} set_property(key, value, ...) -> {success} get_duration(...) -> {duration} get_start(...) -> {start} get_end(...) -> {end} get_source_start_frame(...) -> {frame} get_source_end_frame(...) -> {frame} get_source_start_time(...) -> {time} get_source_end_time(...) -> {time} get_left_offset(...) -> {offset} get_right_offset(...) -> {offset} set_clip_enabled(enabled, ...) -> {success} get_clip_enabled(...) -> {enabled} update_sidecar(...) -> {success} get_unique_id(...) -> {id} get_media_pool_item(...) -> {name, id} get_stereo_convergence(...) -> {values} get_stereo_left_window(...) -> {params} get_stereo_right_window(...) -> {params} get_linked_items(...) -> {items} get_track_type_and_index(...) -> {track_type, track_index} get_source_audio_mapping(...) -> {mapping} load_burnin_preset(name, ...) -> {success} set_name(name, ...) -> {success} get_voice_isolation_state(...) -> {state} set_voice_isolation_state(state, ...) -> {success} get_retime(...) -> {process, motion_estimation} set_retime(process?, motion_estimation?, ...) -> {success} — process: nearest, frame_blend, optical_flow (or 0-3); motion_estimation: 0-6 get_transform(...) -> {Pan, Tilt, ZoomX, ZoomY, RotationAngle, ...} set_transform(Pan?, Tilt?, ZoomX?, ZoomY?, RotationAngle?, AnchorPointX?, AnchorPointY?, Pitch?, Yaw?, FlipX?, FlipY?, ...) -> {success} get_crop(...) -> {CropLeft, CropRight, CropTop, CropBottom, CropSoftness, CropRetain} set_crop(CropLeft?, CropRight?, CropTop?, CropBottom?, CropSoftness?, CropRetain?, ...) -> {success} get_composite(...) -> {Opacity, CompositeMode} set_composite(Opacity?, CompositeMode?, ...) -> {success} get_audio(...) -> {Volume, Pan, AudioSyncOffset, ...} set_audio(Volume?, Pan?, ...) -> {success} get_keyframes(property, ...) -> {property, count, keyframes} add_keyframe(property, frame, value, ...) -> {success} modify_keyframe(property, frame, new_value?, new_frame?, ...) -> {success} delete_keyframe(property, frame, ...) -> {success} set_keyframe_interpolation(property, frame, interpolation, ...) -> {success} — Linear, Bezier, EaseIn, EaseOut, EaseInOut Default: track_type="video", track_index=1, item_index=0

Markers, flags, and clip color on timeline items. Identify by track_type, track_index, item_index (item_index is 0-BASED: 0 = first clip; track_index is 1-based). Actions: add(frame|frame_id|frameId, color?, name?, note?, duration?, custom_data?, ...) -> {success, frame} get_all(...) -> {markers} get_by_custom_data(custom_data, ...) -> {markers} update_custom_data(frame|frame_id|frameId, custom_data, ...) -> {success} get_custom_data(frame|frame_id|frameId, ...) -> {data} delete_by_color(color, ...) -> {success} delete_at_frame(frame|frame_id|frameId, ...) -> {success} delete_by_custom_data(custom_data, ...) -> {success} add_flag(color, ...) -> {success} get_flags(...) -> {flags} clear_flags(color, ...) -> {success} get_clip_color(...) -> {color} set_clip_color(color, ...) -> {success} clear_clip_color(...) -> {success} Default: track_type="video", track_index=1, item_index=0

Fusion composition operations on timeline items. Identify by track_type, track_index, item_index (item_index is 0-BASED: 0 = first clip; track_index is 1-based). Actions: add_comp(...) -> {success} get_comp_count(...) -> {count} get_comp_names(...) -> {names} get_comp_by_name(name, ...) -> {available} get_comp_by_index(index, ...) -> {available} export_comp(path, index, ...) -> {success} import_comp(path, ...) -> {success} delete_comp(name, ...) -> {success} load_comp(name, ...) -> {success} rename_comp(old_name, new_name, ...) -> {success} get_cache_enabled(...) -> {enabled} — Fusion output cache status set_cache(value, ...) -> {success} — value: "Auto", "On", or "Off" Default: track_type="video", track_index=1, item_index=0

Color grading, versions, LUTs, cache, and AI tools on timeline items. Identify by track_type, track_index, item_index (item_index is 0-BASED: 0 = first clip; track_index is 1-based). <when_to_use> - For ANY grade recommendation, lead with action="grade_evidence_base" — composes version_snapshot + node_graph + color_group + coverage_report into one line. - For batch matching, use bulk_match_to_hero (returns confirm_token; dry_run first). - To formalize a recommendation, use propose_grade — validated structured output. - PREFER safe_*/probe_*/*_capabilities/*_boundary_report variants over raw mutators. </when_to_use> <returns> grade_evidence_base -> {evidence_base: str, structured: {coverage, version_snapshot, node_graph, color_group, warnings}} bulk_match_to_hero -> {hero, proposals: [{target_id, name, proposed_cdl|copy_source, warnings}], blocked: [...], confirm_token?} propose_grade -> {accepted: bool, validation, plan_id?, preview_path?, error?} safe_set_cdl -> {success, validation, normalized} safe_copy_grade -> {success, targets, missing} safe_apply_drx -> {success, path, source} # first call may return confirm_token grade_capabilities -> {item_methods, graph_sources, lut_export_types, guards} grade_boundary_report -> {capabilities, item, color_groups, gallery} All actions may return {"error": {code, category, retryable, message, remediation, reason?}}. </returns> PREFER the safe_* / probe_* / *_capabilities / *_boundary_report variants. They validate inputs, support dry_run, and return structured errors. Raw actions are kept for advanced callers but bypass guardrails. Read-only / boundary actions (safe to call any time): grade_evidence_base(include_coverage?, max_nodes?, ...) -> {evidence_base, structured, warnings} PREFERRED PRE-FLIGHT for any color recommendation. Composes version_snapshot + node_graph + color_group + coverage_report into one one-line `evidence_base` summary. Lead your response with the `evidence_base` line. # example: action_help(name='<action_name>') grade_capabilities(...) -> {item_methods, graph_sources, lut_export_types, guards} probe_grade_item(...) -> {methods, versions, node_graph, color_group, cache} probe_node_graph(source?, max_nodes?, ...) -> {available, num_nodes, nodes} grade_version_snapshot(...) -> {current, local, remote} color_group_capabilities(...) -> {count, groups} gallery_capabilities(...) -> {available, still_albums, power_grade_albums} grade_boundary_report(...) -> {capabilities, item, color_groups, gallery} get_current_version(...) -> {version} get_version_names(type?, ...) -> {names} get_node_graph(layer_index?, ...) -> {available} get_color_group(...) -> {name} get_color_cache(...) -> {enabled} get_fusion_cache(...) -> {enabled} Guarded mutators (PREFERRED for grade work): safe_set_cdl(cdl, dry_run?, ...) -> {success, validation, normalized} Validates input, supports dry_run, returns normalized CDL. Use this for primary corrections. # example: action_help(name='<action_name>') safe_copy_grade(target_ids, dry_run?, ...) -> {success, targets, missing} Copies grade to N items; dry_run reports targets without mutating. # example: action_help(name='<action_name>') safe_apply_drx(path, source?, grade_mode?, require_temp_path?) -> {success} REPLACES the target graph. Captures a version snapshot first; require_temp_path defaults True. # example: action_help(name='<action_name>') safe_export_lut(type?, path, require_temp_path?) -> {success, path, size} Sandboxed LUT export. grade_version_restore(name, type?, dry_run?, ...) -> {success} Raw mutators (UNSAFE direct mutation — prefer the safe_* sibling): set_cdl(cdl, ...) -> {success} UNSAFE. No validation; no dry_run. Prefer safe_set_cdl. copy_grades(target_ids, ...) -> {success} UNSAFE. No target existence check. Prefer safe_copy_grade. export_lut(type, path, ...) -> {success} UNSAFE. No path sandboxing. Prefer safe_export_lut. reset_all_node_colors(...) -> {success} DESTRUCTIVE. Erases per-node color labels on the item's graph. add_version(name, type?, ...) -> {success} — type: 0=local, 1=remote load_version(name, type?, ...) -> {success} rename_version(old_name, new_name, type?, ...) -> {success} delete_version(name, type?, ...) -> {success} DESTRUCTIVE. Deletes a named grade version. assign_color_group(group_name, ...) -> {success} remove_from_color_group(...) -> {success} set_color_cache(enabled, ...) -> {success} set_fusion_cache(enabled, ...) -> {success} stabilize(...) -> {success} smart_reframe(...) -> {success} create_magic_mask(mode, ...) -> {success} — mode: "F" forward, "B" backward, "BI" bidirectional regenerate_magic_mask(...) -> {success} Default: track_type="video", track_index=1, item_index=0 For long-form per-action guidance and a worked example, call: timeline_item_color(action="action_help", params={"name": "<action>"})

Take management on timeline items. Identify by track_type, track_index, item_index (item_index is 0-BASED: 0 = first clip; track_index is 1-based). Actions: add(clip_id, start_frame?, end_frame?, ...) -> {success} get_count(...) -> {count} get_selected_index(...) -> {index} get_by_index(index, ...) -> {take} select(index, ...) -> {success} delete(index, ...) -> {success} finalize(...) -> {success} Default: track_type="video", track_index=1, item_index=0

Gallery album management. Actions: get_album_name(album_index?) -> {name} set_album_name(name, album_index?) -> {success} get_current_album() -> {available} set_current_album(album_index) -> {success} get_still_albums() -> {albums} get_power_grade_albums() -> {albums} create_still_album() -> {success} create_power_grade_album() -> {success} album_index is 0-based into the still albums list.

Get media and file listings in a given absolute folder path from Media Storage. Args: folder_path: Absolute path to the folder to list files for.

Manage stills in gallery albums (best results on Color page). Actions: get_stills(album_index?) -> {count} get_label(still_index, album_index?) -> {label} set_label(still_index, label, album_index?) -> {success} import_stills(paths, album_index?) -> {success} export_stills(folder_path, prefix?, format?, album_index?) -> {success} grab_and_export(folder_path, prefix?, format?, album_index?, delete_after?, cleanup?) -> {files} delete_stills(still_indices, album_index?) -> {success} album_index defaults to current album. still_index is 0-based. format for export: dpx, cin, tif, jpg, png, ppm, bmp, xpm, drx (default: dpx). grab_and_export grabs a still from the current frame and exports it immediately, keeping the live GalleryStill reference (more reliable than separate grab + export). Requires Color page. Automatically produces a companion .drx grade file. File data is inlined in the response (DRX as text, images as base64). cleanup (default true) deletes the exported files after inlining. Only files this call produced are ever removed: the export goes to a private staging directory inside folder_path, so anything else written there meanwhile is untouched, and folder_path itself is removed only if this call created it and left it empty. With cleanup false the files are moved up into folder_path without overwriting anything already there.

Node graph operations (color grading nodes). Source can be timeline, timeline item, or color group. PREFER timeline_item_color.probe_node_graph / safe_apply_drx for inspection and replacement. These raw graph actions do not capture a version snapshot before mutating. Read-only: get_num_nodes(source?, ...) -> {count} get_lut(node_index, source?, ...) -> {lut} get_node_cache(node_index, source?, ...) -> {cache} get_node_label(node_index, source?, ...) -> {label} get_tools_in_node(node_index, source?, ...) -> {tools} Mutators (UNSAFE — bypass grade-version snapshot; prefer timeline_item_color.safe_* where available): set_lut(node_index, lut_path, source?, ...) -> {success} UNSAFE. Assigns a LUT to an existing node. No snapshot/dry_run. set_node_cache(node_index, cache_value, source?, ...) -> {success} set_node_enabled(node_index, enabled, source?, ...) -> {success} apply_grade_from_drx(path, grade_mode?, source?, ...) -> {success} CATASTROPHIC. REPLACES the entire node graph. Prefer timeline_item_color.safe_apply_drx (confirm-token gated + validated; it does NOT snapshot — grab a still/.drx backup first, and ALWAYS pass track_type/track_index/item_index explicitly: the default target is video track 1 item 0, not the current clip). All grade_modes are full replacement — there is no append mode. Same-structure applies keep the existing node layout (positions in the .drx are ignored unless the graph differs, e.g. after reset_all_grades). grade_mode: 0="No keyframes" (default), 1="Source Timecode aligned", 2="Start Frames aligned" apply_arri_cdl_lut(source?, ...) -> {success} DESTRUCTIVE. Bakes a CDL LUT into the graph. reset_all_grades(source?, ...) -> {success} CATASTROPHIC. Wipes the entire grade on the source. source: "timeline" (default), "item" (needs track_type/track_index/item_index), "color_group_pre"/"color_group_post" (needs group_name) For long-form per-action guidance and a worked example, call: graph(action="action_help", params={"name": "<action>"})

Manage color groups and their node graphs. Actions: list() -> {groups} get_name(group_name) -> {name} set_name(group_name, new_name) -> {success} get_clips(group_name) -> {clips} get_pre_clip_graph(group_name) -> {available, num_nodes} get_post_clip_graph(group_name) -> {available, num_nodes}

Fusion composition node graph operations. Target comp: - Timeline item: pass clip_id, timeline_item_id, or timeline_item={track_type, track_index, item_index}. Optional comp_name or 1-based comp_index selects a specific comp; otherwise the first comp is used. - Fusion page: omit timeline scope and this uses Fusion().GetCurrentComp(). Use timeline_item_fusion to add/delete/import/export comps on items. Actions: add_tool(tool_type, x?, y?, name?) -> {tool_name, tool_type} delete_tool(tool_name) -> {success} get_tool_list(type?) -> {tools, count} find_tool(name) -> {found, name, type, attrs} connect(target_tool, input_name, source_tool, output_name?) -> {success} disconnect(tool_name, input_name) -> {success} get_inputs(tool_name) -> {inputs} get_outputs(tool_name) -> {outputs} set_input(tool_name, input_name, value, time?) -> {success} get_input(tool_name, input_name, time?) -> {value} set_attrs(tool_name, attrs) -> {success} get_attrs(tool_name) -> {attrs} add_keyframe(tool_name, input_name, time, value) -> {success} get_keyframes(tool_name, input_name) -> {keyframes} delete_keyframe(tool_name, input_name, time) -> {success, time, remaining_keyframes} Deletes on the spline attached to the input. Structured errors when the input is not animated or has no keyframe at that frame. get_comp_info() -> {name, tool_count, attrs} get_position(tool_name) -> {tool_name, x, y} — read a node's FlowView position set_position(tool_name, x, y) -> {success, x, y, readback} — move a node copy_tool(tool_name, name?, x?, y?) -> {success, new_tool, new_tools} — duplicate a node auto_arrange(tool_names?, direction?, spacing?, x?, y?) -> {success, arranged, count} Lay tools out in a row (direction="horizontal", default) or column ("vertical"). set_frame_range(start, end) -> {success} get_frame_range() -> {start, end} — read the comp's render frame range render() -> {success} start_undo(name?) -> {success} end_undo(keep?) -> {success} bulk_set_inputs(ops) -> {results, op_count} — each op requires timeline scope plus tool_name, input_name, value bulk_set_expressions(ops) -> {results, op_count} — each op requires timeline scope plus tool_name, input_name, expression group_settings_export(group_name, path, include_advisory?) -> {path, published_inputs, input_count} group_settings_splice_inputs(source_path, template_path, dest_path?, source_group_name?, template_group_name?) -> {dest_path, summary} Replace a source .setting's `Inputs = ordered() { ... }` block with the matching block from template_path. Both files are read; neither GroupOperator is loaded into Resolve. Inner tools and outer structure are preserved. group_settings_load(group_name, settings_path, backup_path?, undo_name?) -> {settings_path, backup_path} Wraps the LoadSettings call in StartUndo/EndUndo + comp.Lock for reversibility. probe_group_published_inputs(group_name, settings_path?, max_inputs?, time?) -> {live_inputs, file_summary?} fusion_graph_capabilities(...) -> {supported, boundaries, common_tools} probe_fusion_comp(include_io?, max_tools?) -> {name, tool_count, tools} probe_fusion_tool(tool_name, include_io?) -> {found, tool} safe_add_tool(tool_type, name?, dry_run?) -> {success, tool} safe_set_inputs(tool_name, inputs, readback?) -> {success, results} safe_connect_tools(target_tool, input_name, source_tool, dry_run?) -> {success} fusion_boundary_report(include_io?) -> {capabilities, composition} add_fusion_mask(mask_type?, name?, center?|center_x?|center_y?, width?, height?, corner_radius?, angle?, soft_edge?, border_width?, invert?, inputs?, x?, y?, connect_to?, connect_input?, readback?) -> {success, tool_name, inputs_set, connection?} mask_type: "Rectangle" (default) or "Ellipse". One call = add the mask tool, set its params (corner_radius/width/height/center are 0..1), and optionally wire it into connect_to's mask input (connect_input default "EffectMask"). set_text_plus(text, tool_name?, input_name?, readback?) -> {success, tool_name, readback?} Set the text of a Fusion Text+ tool / title template (e.g. a "Deep" Fusion title). Auto-finds the Text+ tool if tool_name is omitted; input_name defaults to "StyledText". get_text_plus(tool_name?, input_name?) -> {tool_name, input_name, text} Common tool_type values: Merge, Background, TextPlus, Transform, Blur, ColorCorrector, RectangleMask, EllipseMask, Tracker, MediaIn, MediaOut, Loader, Saver, Glow, FilmGrain, CornerPositioner, DeltaKeyer, UltraKeyer

Author and install Fusion Fuse plugins (.fuse files). Fuses are Lua plugins (or GLSL view-LUT shaders) that Fusion loads at startup. A NEW Fuse requires a Resolve restart to register; existing Fuses can be edited and reloaded from the Inspector's Edit/Reload buttons without restart. The MCP cannot trigger reload — that's a UI-only action. Actions: path() -> {fuses_dir} list() -> {fuses} — Fuses with the @mcp-fuse marker comment. list(all=true) -> {fuses} — All .fuse files in the directory. install(name, source, overwrite?) -> {success, path} — name: [A-Za-z_][A-Za-z0-9_]* — source: full Fuse source (Lua, or Lua+GLSL for view LUTs) — overwrite: bool (default false) remove(name) -> {success} read(name) -> {source} validate(source, type?) -> {valid, errors, checker} — type: 'lua' (default) | 'glsl' template(kind, name, options?) -> {source, kind, name} — Returns generated source. Pass it to install() to write to disk. — kind: one of the keys returned by list_templates(). See docs/authoring/fuse-dctl-authoring.md for the per-kind option spec. list_templates() -> {kinds}

Author and install DCTL files (Color page custom shaders + ACES transforms). Regular DCTLs live under Resolve's LUT directory and appear as LUT-style entries in the LUT browser, the Clip/Node LUT picker, and the ResolveFX DCTL plugin. After install, call project_settings(action='refresh_luts') to make Resolve pick up the new file. ACES DCTLs (IDT/ODT) live in a separate ACES Transforms directory tree and are scanned only at Resolve startup — install requires a Resolve restart, NOT a LUT refresh. See docs/notes/dctl-notes.md for the full spec and docs/authoring/fuse-dctl-authoring.md for the experimental-tools coverage matrix. Actions: path(category?, subdir?) -> {dctl_dir} — category: 'lut' (default) | 'aces_idt' | 'aces_odt' list(category?, subdir?, all?) -> {files} — Default: only files with @mcp-dctl marker. Pass all=true for everything. install(name, source, category?, subdir?, ext?, overwrite?) -> {success, path} — name: filesystem-safe identifier — source: DCTL source as a string — category: 'lut' (default) | 'aces_idt' | 'aces_odt' — subdir: optional folder under the install root — ext: '.dctl' (default) or '.dctle' (encrypted) remove(name, category?, subdir?, ext?) -> {success} read(name, category?, subdir?, ext?) -> {source, encrypted} validate(source) -> {valid, errors, warnings, checker} template(kind, name, options?) -> {source, kind, name, suggested_category} — kind: 'transform' | 'transform_alpha' | 'transition' | 'matrix' | 'kernel' | 'lut_apply' | 'aces_idt' | 'aces_odt' — `aces_*` kinds set suggested_category to 'aces_idt'/'aces_odt'; pass that as the `category` argument to install(). list_templates() -> {kinds, kind_categories}

Author and install Resolve-page Lua/Python scripts (Workspace → Scripts menu). Scripts live under the per-category subdirs of Resolve's Fusion/Scripts/ directory and appear automatically in Workspace → Scripts → <category> on the page that matches the category. Resolve picks up new scripts without a restart — the menu refreshes the next time it's opened. Categories: 'Edit', 'Color', 'Deliver', 'Comp', 'Tool', 'Utility', 'Views'. 'Utility' shows up everywhere; the rest only on the matching page. Two template kinds: - 'scaffold' minimal stub with Resolve handle setup - 'media_rules' rules-and-variables DSL (declarative VARIABLES + RULES interpreted by an embedded engine; supports sources, extract patterns, transforms, targets, actions, conditions, dry-run, external CSV/JSON data, fuzzy matching, and per-rule metadata) Both languages (Lua and Python) generate fully self-contained scripts. See docs/authoring/script-plugin-authoring.md for the DSL spec. Actions: path(category) -> {scripts_dir} categories() -> {categories} list(category?, all?, language?) -> {scripts} install(name, source, category, language?, overwrite?) -> {success, path} remove(name, category, language) -> {success} read(name, category, language) -> {source} validate(source, language?) -> {valid, errors, checker} template(kind, name, options?) -> {source, kind, name, language} — kind: 'scaffold' | 'media_rules' — options: {language: 'lua'|'py', ...kind-specific} list_templates() -> {kinds} execute(name, category, language, args?, timeout?) -> {success, stdout?, stderr?, exit_code?} — Python: subprocess with stdout/stderr captured. — Lua: fusion.RunScript(); print() output goes to Resolve Console. — args: list of CLI args for the Python subprocess (Python only). — timeout: seconds (default 120 for execute, 60 for run_inline). — Auto-launches Resolve if not running. — Python scripts hard-exit after the script body (guards against fusionscript's segfault-at-exit race), so atexit handlers do not run and non-daemon threads are not joined. Do cleanup inline or in try/finally, not in atexit. run_inline(source, language, timeout?) -> {success, stdout?, stderr?, result?} — Python: writes to temp file with `resolve`/`project`/`mp`/`timeline` pre-bound, runs as subprocess, captures stdout/stderr. — Lua: fusion.Execute(source); return value comes back as `result`. — Use this for ad-hoc one-shot queries without persisting a file. extension_capabilities() -> {paths, templates, lifecycle, safe_guards} probe_fuse_lifecycle(name?, kind?, install?, cleanup?) -> {template, validation, install?, remove?} probe_dctl_lifecycle(name?, kind?, category?, install?, refresh_luts?, cleanup?) -> {template, validation, install?, remove?} probe_script_lifecycle(name?, language?, category?, install?, execute?, cleanup?) -> {template, validation, install?, execute?, remove?} safe_install_extension(extension_type, name, source?|kind?, dry_run?) -> {success} safe_remove_extension(extension_type, name, dry_run?) -> {success} refresh_or_restart_required(extension_type, category?) -> {refresh_luts, restart_required} extension_boundary_report(include_template_matrix?) -> {capabilities, template_matrix, dry_run_probes}

Get list of mounted volumes displayed in Resolve's Media Storage.

Get list of available project presets.

Add specified file/folder paths from Media Storage into current Media Pool folder. Args: file_paths: Simple form — list of absolute file/folder paths. item_infos: Positioned form — list of dicts with keys media (required), startFrame, endFrame. Mirrors MediaStorage.AddItemListToMediaPool([{itemInfo}, ...]) per docs line 210.

Add clip mattes from Media Storage to a MediaPoolItem. Args: media_pool_item_id: The unique ID of the MediaPoolItem. matte_paths: List of absolute file paths for the matte files.

Add timeline mattes from Media Storage to a timeline item. Args: timeline_item_index: 0-based index of the item in the track. matte_paths: List of absolute file paths for the matte files. track_type: Track type ('video' or 'audio'). Default: 'video'. track_index: 1-based track index. Default: 1.

Set a project setting to the specified value. Args: setting_name: The name of the setting to change setting_value: The new value for the setting (can be string, integer, float, or boolean)

Open a project by name. Args: name: The name of the project to open

Create a new project with the given name. Args: name: The name for the new project media_location_path: Optional project media location path (Resolve 20.2.2+).

Save the current project. Note that DaVinci Resolve typically auto-saves projects, so this may not be necessary.

Close the current project. This closes the current project without saving. If you need to save, use the save_project function first.

Set cache mode for the current project. Args: mode: Cache mode to set. Options: 'auto', 'on', 'off'

Set optimized media mode for the current project. Args: mode: Optimized media mode to set. Options: 'auto', 'on', 'off'

Set proxy media mode for the current project. Args: mode: Proxy mode to set. Options: 'auto', 'on', 'off'

Set proxy media quality for the current project. Args: quality: Proxy quality to set. Options: 'quarter', 'half', 'threeQuarter', 'full'

Set cache file path for the current project. Args: path_type: Type of cache path to set. Options: 'local', 'network' path: File system path for the cache

Create a new cloud project. Mirrors ProjectManager.CreateCloudProject({cloudSettings}) per docs lines 138, 576-594. Args: project_name: maps to resolve.CLOUD_SETTING_PROJECT_NAME (required for create). project_media_path: maps to resolve.CLOUD_SETTING_PROJECT_MEDIA_PATH (required for create). is_collab: maps to resolve.CLOUD_SETTING_IS_COLLAB. sync_mode: 'none', 'proxy_only' (default), or 'proxy_and_orig' — maps to resolve.CLOUD_SYNC_*. is_camera_access: maps to resolve.CLOUD_SETTING_IS_CAMERA_ACCESS.

Load a cloud project. Mirrors ProjectManager.LoadCloudProject({cloudSettings}). Per docs line 585, only project_name, project_media_path, and sync_mode are honoured. Args: project_name: required. project_media_path: required. sync_mode: 'none', 'proxy_only', or 'proxy_and_orig'.

Import a cloud project from a file. Mirrors ProjectManager.ImportCloudProject(filePath, {cloudSettings}). Args: file_path: absolute path to the cloud project file (required). project_name, project_media_path, is_collab, sync_mode, is_camera_access: cloudSettings dict keys; same semantics as create_cloud_project_tool.

Get a timeline by its 1-based index. Args: index: 1-based timeline index.

Restore a cloud project from a folder. Mirrors ProjectManager.RestoreCloudProject(folderPath, {cloudSettings}). Args: folder_path: absolute path to the cloud project folder (required). project_name, project_media_path, is_collab, sync_mode, is_camera_access: cloudSettings dict keys; same semantics as create_cloud_project_tool.

Set a project property value. Args: property_name: Name of the property to set property_value: Value to set for the property

Set timeline format (resolution and frame rate). Args: width: Timeline width in pixels height: Timeline height in pixels frame_rate: Timeline frame rate interlaced: Whether the timeline should use interlaced processing

Set SuperScale settings for the current project. Args: enabled: Whether SuperScale is enabled quality: SuperScale quality (0=Auto, 1=Better Quality, 2=Smoother)

Set color science mode for the current project. Args: mode: Color science mode ('YRGB', 'YRGB Color Managed', 'ACEScct', or numeric value)

Set timeline color space and gamma. Args: color_space: Timeline color space (e.g., 'Rec.709', 'DCI-P3 D65', 'Rec.2020') gamma: Timeline gamma (e.g., 'Rec.709 Gamma', 'Gamma 2.4')

Archive a project to a file with optional media. Args: project_name: Name of the project to archive. archive_path: Absolute path for the archive file (.dra). archive_src_media: Include source media in archive. Default: True. archive_render_cache: Include render cache. Default: True. archive_proxy_media: Include proxy media. Default: False.

Delete a project from the current database. WARNING: This is irreversible. Args: project_name: Name of the project to delete.

Create a new folder in the current project folder location. Args: folder_name: Name of the folder to create.

Delete a folder from the current project folder location. Args: folder_name: Name of the folder to delete.

Get list of folders in the current project folder location.

Navigate to the root project folder.

Navigate up one level in the project folder hierarchy.

Get the name of the current project folder.

Open/navigate into a project folder. Args: folder_name: Name of the folder to open.

Import a project from a .drp file. Args: file_path: Absolute path to the .drp project file.

Export a project to a .drp file. Args: project_name: Name of the project to export. file_path: Absolute path for the exported .drp file. with_stills_and_luts: Include stills and LUTs in export. Default: True.

Restore a project from an archive (.dra) file. Args: file_path: Absolute path to the .dra archive file.

Get information about the current database.

Get list of all available databases.

Switch to a different database. Args: db_info: Database info dict with keys 'DbType' and 'DbName'. Example: {"DbType": "Disk", "DbName": "Local Database"}

Apply a project preset to the current project. Args: preset_name: Name of the preset to apply.

Delete a specific render job by its ID. Args: job_id: The unique ID of the render job to delete.

Get list of all render jobs in the queue.

Start rendering jobs. If no job IDs specified, renders all queued jobs. Args: job_ids: Optional list of job IDs to render. If None, renders all. is_interactive_mode: If True, enables interactive rendering mode.

Stop the current rendering process.

Check if rendering is currently in progress.

Load a render preset by name. Args: preset_name: Name of the render preset to load.

Save current render settings as a new preset. Args: preset_name: Name for the new render preset.

Delete a render preset. Args: preset_name: Name of the render preset to delete.

Set render settings for the current project. Mirrors Project.SetRenderSettings({settings}) per docs lines 765-799. Args: settings: Dict of render settings. All documented keys are forwarded: SelectAllFrames (bool), MarkIn (int), MarkOut (int), TargetDir (str), CustomName (str), UniqueFilenameStyle (0=Prefix/1=Suffix), ExportVideo (bool), ExportAudio (bool), FormatWidth (int), FormatHeight (int), FrameRate (float), PixelAspectRatio (float), VideoQuality (int/str), AudioCodec (str), AudioBitDepth (int), AudioSampleRate (int), ColorSpaceTag (str), GammaTag (str), ExportAlpha (bool), EncodingProfile (str), MultiPassEncode (bool), AlphaMode (int), NetworkOptimization (bool), ClipStartFrame (int), TimelineStartTimecode (str), ReplaceExistingFilesInPlace (bool), ExportSubtitle (bool), SubtitleFormat ("BurnIn", "EmbeddedCaptions", "SeparateFile").

Get the status of a specific render job. Args: job_id: The unique ID of the render job.

Get all available render formats.

Get available codecs for a given render format. Args: format_name: Render format id or display name (e.g. 'mov', 'QuickTime').

Get the current render format and codec setting.

Set the render format and codec. Args: format_name: Render format id or display name (e.g. 'mov', 'QuickTime'). codec_name: Codec id or display name (e.g. 'ProRes422HQ', 'Apple ProRes 422 HQ', 'H.264').

Get the current render mode (0=Individual Clips, 1=Single Clip).

Set the render mode. Args: mode: 0 for Individual Clips, 1 for Single Clip.

Get available render resolutions for a format/codec combination. Args: format_name: Render format (e.g. 'mp4'). codec_name: Codec name (e.g. 'H264').

Refresh the LUT list in the project. Call after adding new LUT files.

Get the unique ID of the current project.

Insert audio file to current track at playhead position. Args: file_path: Absolute path to the audio file.

Load a burn-in preset by name for the project. Args: preset_name: Name of the burn-in preset to load.

Export the current frame as a still image. Args: file_path: Absolute path for the exported still image.

Get list of all color groups in the current project.

Create a new color group in the current project. Args: group_name: Name for the new color group.

Delete a color group from the current project. Args: group_name: Name of the color group to delete.

Rename a color group. Mirrors ColorGroup.SetName(groupName). Args: group_name: Current name of the color group. new_name: New name to assign.

Apply a Fairlight preset to the current timeline. Args: preset_name: Name of the Fairlight preset to apply.

Get list of available quick export render presets.

Render the current timeline using a Quick Export preset. Mirrors Project.RenderWithQuickExport(preset_name, {param_dict}) per docs line 179. Args: preset_name: Name of the Quick Export preset (from get_quick_export_render_presets). target_dir: Output directory; maps to TargetDir in param_dict. custom_name: Output filename; maps to CustomName. video_quality: Video quality setting (int or string per render-settings spec); maps to VideoQuality. enable_upload: Enable direct upload for supported web presets; maps to EnableUpload.

Add a render job based on current render settings to the render queue. Returns the unique job ID string for the new render job. Configure render settings first with set_render_settings, set_render_format_and_codec, etc.

Load a cloud project from DaVinci Resolve cloud. Args: project_name: Name of the cloud project to load. project_media_path: Local path for project media cache. sync_mode: Sync mode - 'proxy' or 'full' (default: 'proxy').

Generate AI text-to-speech audio and add it to the media pool (Resolve 21+). Requires the AI Speech Generator Extra. Creates a NEW audio MediaPoolItem; if add_to_timeline is True it is placed on the timeline at the given timecode. Args: text_input: Text to synthesize (required). voice_model: Voice model name (e.g. "Female 1", "Male 1", "Custom Voice"). timecode: Timeline timecode to place the clip at when add_to_timeline is True. add_to_timeline: Whether to add the generated clip to the timeline. audio_track: Audio track index for timeline placement. custom_voice_file: Full path to a custom voice file (for "Custom Voice"). speed: Speech speed. variation: Voice variation. pitch: Voice pitch. generation_id: Generation ID. filename: Output filename.

Get per-project attributes for every project in the current folder (Resolve 21.0.4+). Calls ProjectManager.GetProjectAttributesInCurrentFolder(). Returns a dict keyed by project name with 'lastModifiedDate', 'creationDate', 'notes' and 'liveCollaborationMode' — without loading any project.

Get the name of the current gallery album.

Set the name of the current gallery album. Args: name: New album name.

Get list of all gallery still albums.

Get list of all gallery power grade albums.

Get the current still album.

Set the current still album by index. Args: album_index: 0-based index of the album in GetGalleryStillAlbums() list.

Create a new gallery still album. Args: album_name: Optional name for the new album.

Create a new gallery power grade album. Args: album_name: Optional name for the new album.

Get list of stills in a gallery album. Args: album_index: 0-based index of the album. Default: 0.

Open the Project Settings dialog in DaVinci Resolve.

Get the label of a still in a gallery album. Args: album_index: 0-based album index. still_index: 0-based still index.

Set the label of a still in a gallery album. Args: album_index: 0-based album index. still_index: 0-based still index. label: New label for the still.

Import stills from file paths into a gallery album. Args: album_index: 0-based album index. file_paths: List of absolute file paths to import.

Export stills from a gallery album. Args: album_index: 0-based album index. folder_path: Directory to export to. file_prefix: Filename prefix. Default: 'still'. format: File format (dpx, cin, tif, jpg, png, ppm, bmp, xpm, drx). Default: 'dpx'.

Delete stills from a gallery album. Args: album_index: 0-based album index. still_indices: List of 0-based still indices to delete.

Get number of nodes in the color graph for a timeline item. Args: item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Set LUT on a node in the color graph. Args: node_index: 1-based node index. lut_path: Absolute or relative LUT path. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get LUT path on a node in the color graph. Args: node_index: 1-based node index. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Set the cache mode on a node. Args: node_index: 1-based node index. cache_value: -1=Auto, 0=Disabled, 1=Enabled. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get the cache mode of a node. Args: node_index: 1-based node index. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get the label of a node. Args: node_index: 1-based node index. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get list of tools used in a node. Args: node_index: 1-based node index. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Enable or disable a node. Args: node_index: 1-based node index. is_enabled: True to enable, False to disable. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Apply a grade from a .drx file to a timeline item's graph. Args: drx_path: Absolute path to the .drx file. grade_mode: 0=No keyframes, 1=Source Timecode aligned, 2=Start Frames aligned. item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Apply ARRI CDL and LUT to a timeline item's graph. Args: item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Reset all grades on a timeline item's graph. Args: item_index: 0-based timeline item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get clips in a color group for the current timeline. Args: group_name: Name of the color group.

Get the pre-clip node graph for a color group. Args: group_name: Name of the color group.

Get the post-clip node graph for a color group. Args: group_name: Name of the color group.

Import media into the current project's media pool. Args: paths: Simple form — list of file or folder paths to import. clip_infos: Image-sequence form — list of dicts with keys FilePath (required), StartIndex, EndIndex. Mirrors MediaPool.ImportMedia([{clipInfo}, ...]). Each entry imports as one MediaPoolItem unless 'Show Individual Frames' is enabled. Example: [{"FilePath": "frame_%03d.dpx", "StartIndex": 1, "EndIndex": 100}] file_path: Single-path convenience for backward compatibility.

Append clips to the current timeline. Mirrors MediaPool.AppendToTimeline per docs lines 219-221. Two forms: Args: clip_ids: Simple form — list of MediaPoolItem unique IDs to append in order. clip_infos: Positioned form — list of dicts with keys clip_id (or media_pool_item_id), start_frame, end_frame, record_frame, track_index, and optional media_type (1=video only, 2=audio only). record_frame is relative to the current timeline start frame by default; pass record_frame_mode="absolute" for raw Resolve recordFrame values. Returns timeline_item_id per appended item.

Create a stacked multicam prep timeline from Media Pool clips. This does not create a native Resolve multicam clip because the public scripting API does not expose native multicam creation. It creates a prep timeline with one angle per video track and optional matching audio tracks. Args: name: Name for the setup timeline. clip_ids: Simple list of MediaPoolItem unique IDs, one per angle. angles: Detailed angle rows with clip_id, optional angle_name, start_frame/end_frame, record_frame, source_timecode, track_index. sync_mode: 'stack_start', 'record_frame', or 'source_timecode'. include_audio: Also append audio-only rows. audio_track_mode: 'matching' for per-angle audio tracks, or 'first'. start_timecode: Optional timeline start timecode to set after creation. timeline_start_timecode: Base timecode for source_timecode sync math. record_frame_start: Timeline-relative start frame for stack_start mode. frame_rate: Fallback frame rate for duration/timecode parsing. audio_type: Resolve audio track type when tracks must be added. dry_run: Return the plan without creating a timeline.

Sync audio across multiple clips. Mirrors MediaPool.AutoSyncAudio([items], {audioSyncSettings}) per docs lines 600-614. Args: clip_ids: List of MediaPoolItem unique IDs to sync. sync_mode: 'waveform' or 'timecode' (default on Resolve side: 'timecode'). channel_number: int >= 1 for channel offset, or 'automatic' (-1) / 'mix' (-2). Only used in waveform mode. retain_embedded_audio: keep clip's embedded audio after sync. retain_video_metadata: keep video clip's metadata after sync.

Create a new subfolder in the current Media Pool folder. Args: folder_name: Name of the subfolder to create.

Refresh all folders in the Media Pool.

Import a timeline from a file (AAF, EDL, XML, FCPXML, DRT, ADL, OTIO). Args: file_path: Absolute path to the timeline file. import_options: Optional dict of import options.

Delete timelines by their unique IDs. Args: timeline_ids: List of timeline unique IDs to delete.

Open the Preferences dialog in DaVinci Resolve.

Navigate to a specific folder in the Media Pool. Args: folder_path: Folder path using '/' separators from root (e.g. 'Master/Footage').

Delete clips from the Media Pool by their unique IDs. Args: clip_ids: List of clip unique IDs to delete.

Import a Media Pool folder structure from a file. Args: file_path: Absolute path to the file.

Delete folders from the current Media Pool location. Args: folder_names: List of folder names to delete.

Move clips to a different Media Pool folder. Args: clip_ids: List of clip unique IDs to move. target_folder_path: Path to target folder (e.g. 'Master/Footage').

Move folders to a different Media Pool location. Args: folder_names: List of folder names to move. target_folder_path: Path to target folder.

Get list of clip mattes for a MediaPoolItem. Args: clip_id: Unique ID of the clip.

Get list of timeline mattes for a timeline item. Args: item_index: 0-based index of the item in the track. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Delete clip mattes from a MediaPoolItem. Args: clip_id: Unique ID of the clip. matte_paths: List of matte file paths to delete.

Export metadata of clips in the current Media Pool folder to a CSV file. Args: file_path: Absolute path for the exported CSV file.

Get the unique ID of the Media Pool.

Create a stereo clip from left and right eye clips. Args: left_clip_id: Unique ID of the left eye clip. right_clip_id: Unique ID of the right eye clip.

Get currently selected clips in the Media Pool.

Set a clip as selected in the Media Pool. Args: clip_id: Unique ID of the clip to select.

Create a new timeline with the given name. Args: name: The name for the new timeline

Switch to a timeline by name. Args: name: The name of the timeline to set as current

List all timelines in the current project as a tool.

Rename the current timeline. Args: name: New name for the timeline.

Set the start timecode of the current timeline. Args: timecode: Timecode string (e.g. '01:00:00:00').

Get the current playhead timecode.

Set the playhead to a specific timecode. Args: timecode: Timecode string (e.g. '01:00:05:00').

Add a new track to the timeline. Mirrors Timeline.AddTrack(trackType, {newTrackOptions}) per docs line 327. Args: track_type: 'video', 'audio', or 'subtitle'. audio_type: For audio tracks: 'mono', 'stereo', 'lrc', 'lcr', 'lrcs', 'lcrs', 'quad', '5.0', '5.0film', '5.1', '5.1film', '7.0', '7.0film', '7.1', '7.1film', 'adaptive1' through 'adaptive36'. Defaults to 'mono' for audio tracks if omitted. index: 1-based track index where 1 <= index <= GetTrackCount(track_type) + 1. If omitted or out of bounds, the track is appended.

Delete a track from the timeline. Args: track_type: 'video', 'audio', or 'subtitle'. track_index: 1-based track index to delete.

Get the sub-type of a track (e.g. mono, stereo for audio). Args: track_type: 'video' or 'audio'. track_index: 1-based track index.

Enable or disable a track. Args: track_type: 'video' or 'audio'. track_index: 1-based track index. enabled: True to enable, False to disable.

Check if a track is enabled. Args: track_type: 'video' or 'audio'. track_index: 1-based track index.

Lock or unlock a track. Args: track_type: 'video' or 'audio'. track_index: 1-based track index. locked: True to lock, False to unlock.

Check if a track is locked. Args: track_type: 'video' or 'audio'. track_index: 1-based track index.

Get voice isolation state for an audio track. Args: track_index: 1-based audio track index.

Set voice isolation state for an audio track. Args: track_index: 1-based audio track index. state: Dict with isEnabled (bool) and amount (0-100).

Delete clips from the timeline. Args: clip_ids: List of clip unique IDs to delete. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Link or unlink clips in the timeline. Args: clip_ids: List of clip unique IDs. linked: True to link, False to unlink. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Add a marker to the current timeline. Args: frame_id: Frame number for the marker, relative to the timeline start (frame 0 = first frame, even when the timeline starts at 01:00:00:00). color: Marker color (Blue, Cyan, Green, Yellow, Red, Pink, Purple, Fuchsia, Rose, Lavender, Sky, Mint, Lemon, Sand, Cocoa, Cream). name: Marker name. note: Marker note. Default: empty. duration: Duration in frames. Default: 1. custom_data: Custom data string. Default: empty.

Get all markers on the current timeline.

Find a timeline marker by its custom data. Args: custom_data: Custom data string to search for.

Update the custom data of a timeline marker. Args: frame_id: Frame number of the marker. custom_data: New custom data string.

Get the custom data of a timeline marker. Args: frame_id: Frame number of the marker.

Delete all timeline markers of a specific color. Args: color: Color of markers to delete. Use '' to delete all.

Delete a timeline marker at a specific frame. Args: frame_id: Frame number of the marker.

Delete a timeline marker by custom data. Args: custom_data: Custom data of the marker to delete.

Get the name of a track. Args: track_type: 'video' or 'audio'. track_index: 1-based track index.

Set the name of a track. Args: track_type: 'video' or 'audio'. track_index: 1-based track index. name: New track name.

Duplicate the current timeline.

Create a compound clip from selected items. Mirrors Timeline.CreateCompoundClip([timelineItems], {clipInfo}) per docs line 369. Args: clip_ids: List of timeline item unique IDs. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1. name: Optional name for the compound clip; maps to clipInfo.name. start_timecode: Optional start timecode (e.g. "01:00:00:00"); maps to clipInfo.startTimecode.

Create a Fusion clip from selected items. Args: clip_ids: List of timeline item unique IDs. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Import content into the current timeline from a file. Args: file_path: Path to the file to import (AAF, EDL, XML, etc.). import_options: Optional dict of import options.

Export the current timeline to a file. Args: file_path: Output file path. export_type: Export type (e.g. 'EXPORT_AAF', 'EXPORT_EDL', 'EXPORT_FCP_7_XML', 'EXPORT_FCPXML_1_10', 'EXPORT_DRT', 'EXPORT_TEXT_CSV', 'EXPORT_TEXT_TAB', 'EXPORT_OTIO', 'EXPORT_ALE'). export_subtype: Export subtype for AAF/EDL. Default: 'EXPORT_NONE'.

Insert a generator into the timeline. Args: generator_name: Name of the generator to insert. duration: Optional duration in frames.

Insert a Fusion generator into the timeline. Args: generator_name: Name of the Fusion generator.

Insert a Fusion composition into the timeline.

Insert an OFX generator into the timeline. Args: generator_name: Name of the OFX generator.

Insert a title into the timeline. Args: title_name: Name of the title to insert.

Insert a Fusion title into the timeline. Args: title_name: Name of the Fusion title.

Grab a still from the current frame of the timeline.

Grab stills from all frames at the current position across all timelines.

Get the unique ID of the current timeline.

Create subtitles from audio in the current timeline. Mirrors Timeline.CreateSubtitlesFromAudio({autoCaptionSettings}) per docs lines 718-761. Args: language: 'auto', 'danish', 'dutch', 'english', 'french', 'german', 'italian', 'japanese', 'korean', 'mandarin_simplified', 'mandarin_traditional', 'norwegian', 'portuguese', 'russian', 'spanish', 'swedish'. Default: 'auto'. preset: 'default', 'teletext', 'netflix'. Default: 'default'. chars_per_line: Integer 1-60. Resolve default is 42 (or 16 for Netflix preset). line_break: 'single' or 'double'. Default: 'single' on Resolve side. gap: Integer 0-10. Default: 0 on Resolve side.

Detect scene cuts in the current timeline.

Convert the current timeline to stereo.

Get the node graph for the current timeline.

Analyze Dolby Vision for the current timeline.

Get the current video item at the playhead.

Get DaVinci Resolve version as structured fields [major, minor, patch, build, suffix].

Get thumbnail image data for the current clip. Switches to the Color page for the read (GetCurrentClipThumbnailImage only returns data there) and restores the previous page after. Args: width: Thumbnail width. Default: 320. height: Thumbnail height. Default: 180.

Get the MediaPoolItem for the current timeline.

Get mark in/out points for the current timeline.

Set mark in/out points for the current timeline. Args: mark_in: Mark in frame number. mark_out: Mark out frame number.

Clear mark in/out points for the current timeline.

Create a new timeline from specified media pool clips. Args: name: Name for the new timeline. clip_ids: Simple form — list of MediaPoolItem unique IDs to append end-to-end. clip_infos: Positioned form — list of dicts with keys clip_id (or media_pool_item_id), start_frame, end_frame, record_frame. record_frame is relative to the created timeline start frame by default; pass record_frame_mode="absolute" for raw Resolve values. If both are None, uses the currently selected media pool clips.

Set a timeline setting value. Args: setting_name: Name of the timeline setting to set (e.g. 'useCustomSettings', 'timelineFrameRate', 'timelineResolutionWidth', 'timelineResolutionHeight', 'timelineOutputResolutionWidth', 'timelineOutputResolutionHeight', 'colorSpaceTimeline', 'colorSpaceOutput'). setting_value: Value to set for the setting (string).

Get the timeline items currently selected in the timeline (Resolve 21.0.4+). Calls Timeline.GetSelectedClips() on the current timeline. An empty list means nothing is selected — that is an answer, not a failure. (Distinct from get_selected_clips, which reads the Media Pool selection.)

Switch to a specific page in DaVinci Resolve. Args: page: The page to switch to. Options: 'media', 'cut', 'edit', 'fusion', 'color', 'fairlight', 'deliver'

Get human-readable help for a DaVinci Resolve API object. Args: object_type: Type of object to get help for ('resolve', 'project_manager', 'project', 'media_pool', 'timeline', 'media_storage')

Inspect a custom DaVinci Resolve API object by path. Args: object_path: Path to the object using dot notation (e.g., 'resolve.GetMediaStorage()')

Save the current UI layout as a preset. Calls Resolve.SaveLayoutPreset() to save the current UI layout. Args: preset_name: Name for the saved preset.

Load a UI layout preset. Calls Resolve.LoadLayoutPreset() to load a saved UI layout. Args: preset_name: Name of the preset to load.

Export a layout preset to a file. Calls Resolve.ExportLayoutPreset() to export a preset to disk. Args: preset_name: Name of the preset to export. export_path: Absolute file path to export the preset to.

Import a layout preset from a file. Calls Resolve.ImportLayoutPreset() to import a preset from disk. Args: import_path: Absolute path to the preset file to import. preset_name: Name to save the imported preset as (uses filename if None).

Delete a layout preset. Calls Resolve.DeleteLayoutPreset() to remove a saved preset. Args: preset_name: Name of the preset to delete.

Disable all background tasks for the current Resolve session (Resolve 21+). Useful before heavy scripted operations so Resolve does not run background work that competes for resources. Resets when Resolve restarts.

Quit DaVinci Resolve application. Args: force: Whether to force quit even if unsaved changes (potentially dangerous) save_project: Whether to save the project before quitting

Restart DaVinci Resolve application. Args: wait_seconds: Seconds to wait between quit and restart

Get the Fusion object. Starting point for Fusion scripts.

Overwrite an existing layout preset with the current UI layout. Args: preset_name: Name of the preset to overwrite.

Import a render preset from a file. Args: preset_path: Absolute path to the render preset file.

Export a render preset to a file. Args: preset_name: Name of the render preset to export. export_path: Absolute path where the preset file will be saved.

Import a burn-in preset from a file. Args: preset_path: Absolute path to the burn-in preset file.

Export a burn-in preset to a file. Args: preset_name: Name of the burn-in preset to export. export_path: Absolute path where the preset file will be saved.

Get the current keyframe mode in Resolve. Returns 0=ALL, 1=COLOR, 2=SIZING.

Set the keyframe mode in Resolve. Args: mode: Keyframe mode - 0=All, 1=Color, 2=Sizing.

Get the available Fairlight preset names.

Quit DaVinci Resolve. WARNING: This will close the application.

Get the names of all saved UI layout presets (Resolve 21.0.4+). Calls Resolve.GetLayoutPresetList(). These are the names accepted by load_layout_preset_tool, update_layout_preset, export_layout_preset_tool and delete_layout_preset_tool.

Get the names of all saved data burn-in presets (Resolve 21.0.4+). Calls Resolve.GetBurnInPresetList(). These are the names accepted by the 'DataBurnIn' render setting, project load_burn_in_preset, and export_burn_in_preset.

Delete a data burn-in preset by name (Resolve 21.0.4+). Args: preset_name: Name of the burn-in preset to delete.

Get the names of all saved user-preferences presets (Resolve 21.0.4+). Calls Resolve.GetUserPreferencesPresetList().

Save the current user preferences as a named preset (Resolve 21.0.4+). Args: preset_name: Name for the new user-preferences preset.

Load a user-preferences preset (Resolve 21.0.4+). SESSION-WIDE: this swaps the user's global Resolve preferences, not a project setting. It affects every project open in this Resolve instance. Only call when the user explicitly asked for the switch. Args: preset_name: Name of the user-preferences preset to load.

Delete a user-preferences preset by name (Resolve 21.0.4+). Args: preset_name: Name of the user-preferences preset to delete.

Import a user-preferences preset from a file (Resolve 21.0.4+). The imported preset is NOT auto-loaded; it takes its name from the file when preset_name is omitted (measured on Studio 21.0.4.5). Follow with load_user_preferences_preset to activate it. Args: import_path: Absolute path to the preset file to import. preset_name: Name to save the imported preset as (filename if None).

Export a user-preferences preset to a file (Resolve 21.0.4+). Args: preset_name: Name of the user-preferences preset to export. export_path: Absolute path where the preset file will be saved.

Export a folder to a DRB file or other format. Args: folder_name: Name of the folder to export export_path: Path to save the exported file export_type: Export format (DRB is default and currently the only supported option)

Transcribe audio for all clips in a folder. The transcription language follows the project's Speech Recognition setting (Project Settings > Subtitles), not a per-call argument. Args: folder_name: Name of the folder containing clips to transcribe use_speaker_detection: Optional (Resolve 21+). Enable speaker detection. If omitted, the project's setting is used.

Clear audio transcription for all clips in a folder. Args: folder_name: Name of the folder to clear transcriptions from

Get list of clips in a Media Pool folder. Args: folder_path: Path from root. Empty for current folder.

Get list of subfolders in a Media Pool folder. Args: folder_path: Path from root. Empty for current folder.

Check if a Media Pool folder is stale (needs refresh). Args: folder_path: Path from root. Empty for current folder.

Get the unique ID of a Media Pool folder. Args: folder_path: Path from root. Empty for current folder.

Export a Media Pool folder to a file. Args: file_path: Absolute path for the export. folder_path: Path from root. Empty for current folder.

Transcribe audio for all clips in a Media Pool folder. Args: folder_path: Path from root. Empty for current folder.

Clear transcription for all clips in a Media Pool folder. Args: folder_path: Path from root. Empty for current folder.

Classify audio of all clips in a Media Pool folder into categories (Resolve 21+). Args: folder_path: Path from root. Empty for current folder.

Clear audio classification for all clips in a Media Pool folder (Resolve 21+). Args: folder_path: Path from root. Empty for current folder.

Run IntelliSearch analysis on all clips in a folder (Resolve 21+, requires AI IntelliSearch Extra). Args: folder_path: Path from root. Empty for current folder. identify_faces: Whether to identify faces. is_better_mode: Use Better mode (vs Faster).

Run Slate analysis on all clips in a folder (Resolve 21+, requires AI Slate ID Extra). Args: folder_path: Path from root. Empty for current folder. marker_color: Marker color for detected slates (Blue, Cyan, Green, Yellow, Red, Pink, Purple, Fuchsia, Rose, Lavender, Sky, Mint, Lemon, Sand, Cocoa, Cream).

Render motion-deblurred copies of all clips in a folder (Resolve 21+). Creates NEW media files; source media is not modified. Args: folder_path: Path from root. Empty for current folder. deblur_option: Settings dict (FileName, Format, Codec, EncodingProfile, UseExtremeMode, UseMarkInMarkOut, RenderAtSourceRes, UseMoreGpuMemory).

Set a transform property for a timeline item. Args: timeline_item_id: The ID of the timeline item to modify property_name: The name of the property to set. Options include: 'Pan', 'Tilt', 'ZoomX', 'ZoomY', 'Rotation', 'AnchorPointX', 'AnchorPointY', 'Pitch', 'Yaw' property_value: The value to set for the property

Set a crop property for a timeline item. Args: timeline_item_id: The ID of the timeline item to modify crop_type: The type of crop to set. Options: 'Left', 'Right', 'Top', 'Bottom' crop_value: The value to set for the crop (typically 0.0 to 1.0)

Set composite properties for a timeline item. Args: timeline_item_id: The ID of the timeline item to modify composite_mode: Optional composite mode to set (e.g., 'Normal', 'Add', 'Multiply') opacity: Optional opacity value to set (0.0 to 1.0)

Set retiming properties for a timeline item. Args: timeline_item_id: The ID of the timeline item to modify speed: Optional speed factor (e.g., 0.5 for 50%, 2.0 for 200%) process: Optional retime process. Options: 'NearestFrame', 'FrameBlend', 'OpticalFlow'

Set stabilization properties for a timeline item. Args: timeline_item_id: The ID of the timeline item to modify enabled: Optional boolean to enable/disable stabilization method: Optional stabilization method. Options: 'Perspective', 'Similarity', 'Translation' strength: Optional strength value (0.0 to 1.0)

Set audio properties for a timeline item. Args: timeline_item_id: The ID of the timeline item to modify volume: Optional volume level (usually 0.0 to 2.0, where 1.0 is unity gain) pan: Optional pan value (-1.0 to 1.0, where -1.0 is left, 0 is center, 1.0 is right) eq_enabled: Optional boolean to enable/disable EQ

Add a keyframe at the specified frame for a timeline item property. Args: timeline_item_id: The ID of the timeline item to add keyframe to property_name: The name of the property to keyframe (e.g., 'Pan', 'ZoomX') frame: Frame position for the keyframe value: Value to set at the keyframe

Modify an existing keyframe by changing its value or frame position. Args: timeline_item_id: The ID of the timeline item property_name: The name of the property with keyframe frame: Current frame position of the keyframe to modify new_value: Optional new value for the keyframe new_frame: Optional new frame position for the keyframe

Delete a keyframe at the specified frame for a timeline item property. Args: timeline_item_id: The ID of the timeline item property_name: The name of the property with keyframe to delete frame: Frame position of the keyframe to delete

Set the interpolation type for a keyframe. Args: timeline_item_id: The ID of the timeline item property_name: The name of the property with keyframe frame: Frame position of the keyframe interpolation_type: Type of interpolation. Options: 'Linear', 'Bezier', 'Ease-In', 'Ease-Out'

Enable keyframe mode for a timeline item. Args: timeline_item_id: The ID of the timeline item keyframe_mode: Keyframe mode to enable. Options: 'All', 'Color', 'Sizing'

Get comprehensive info about a timeline item. Args: item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Rename a timeline item. Args: name: New timeline item name. item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get source start time of a timeline item. Args: item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Set a property on a timeline item. Args: property_name: Property name (Pan, Tilt, ZoomX, ZoomY, RotationAngle, Opacity, CropLeft, CropRight, CropTop, CropBottom, etc.). property_value: Value to set. item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get property of a timeline item. Args: property_name: Property name, or empty for all. Default: ''. item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Export a Fusion composition to file. Args: file_path: Output path for the .comp file. comp_index: 1-based Fusion comp index. Default: 1. item_index: 0-based item index. Default: 0.

Add a marker to a timeline item. Args: frame_id: Frame offset within the item. color: Marker color. name: Marker name. note: Marker note. Default: ''. duration: Duration in frames. Default: 1. custom_data: Custom data. Default: ''. item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get all markers on a timeline item. Args: item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Delete markers by color on a timeline item. Args: color: Color to delete. '' for all. item_index: 0-based item index. Default: 0.

Delete a marker at a frame on a timeline item. Args: frame_id: Frame number. item_index: 0-based item index. Default: 0.

Delete a marker by custom data on a timeline item. Args: custom_data: Custom data of the marker. item_index: 0-based item index. Default: 0.

Find marker by custom data. Args: custom_data: Custom data to search for. item_index: 0-based item index. Default: 0.

Update marker custom data. Args: frame_id: Frame number. custom_data: New custom data. item_index: 0-based item index. Default: 0.

Get marker custom data. Args: frame_id: Frame number. item_index: 0-based item index. Default: 0.

Add a flag to a timeline item. Args: color: Flag color. item_index: 0-based item index. Default: 0.

Get flags on a timeline item. Args: item_index: 0-based item index. Default: 0.

Clear flags from a timeline item. Args: color: Color to clear, or '' for all. item_index: 0-based item index. Default: 0.

Get clip color of a timeline item. Args: item_index: 0-based item index. Default: 0.

Set clip color of a timeline item. Args: color: Color name. item_index: 0-based item index. Default: 0.

Clear clip color from a timeline item. Args: item_index: 0-based item index. Default: 0.

Add a new Fusion composition to a timeline item. Args: item_index: 0-based item index. Default: 0.

Import a Fusion composition from file. Args: file_path: Path to the .comp file. item_index: 0-based item index. Default: 0.

Get the media ID for a clip. Args: clip_id: Unique ID of the clip.

Delete a Fusion composition by name. Args: comp_name: Name of the Fusion composition. item_index: 0-based item index. Default: 0.

Load a Fusion composition by name. Args: comp_name: Name of the Fusion composition. item_index: 0-based item index. Default: 0.

Rename a Fusion composition. Args: old_name: Current name. new_name: New name. item_index: 0-based item index. Default: 0.

Get Fusion composition info for a timeline item. Args: item_index: 0-based item index. Default: 0.

Add a new color version to a timeline item. Args: version_name: Name for the new version. version_type: 0=Local, 1=Remote. Default: 0. item_index: 0-based item index. Default: 0.

Get the current color version of a timeline item. Args: item_index: 0-based item index. Default: 0.

Delete a color version. Args: version_name: Name of the version. version_type: 0=Local, 1=Remote. Default: 0. item_index: 0-based item index. Default: 0.

Load a color version. Args: version_name: Name of the version. version_type: 0=Local, 1=Remote. Default: 0. item_index: 0-based item index. Default: 0.

Rename a color version. Args: old_name: Current version name. new_name: New version name. version_type: 0=Local, 1=Remote. Default: 0. item_index: 0-based item index. Default: 0.

Get list of version names. Args: version_type: 0=Local, 1=Remote. Default: 0. item_index: 0-based item index. Default: 0.

Set CDL (Color Decision List) values on a timeline item. Args: cdl: Dict with CDL values: {'NodeIndex': str, 'Slope': str, 'Offset': str, 'Power': str, 'Saturation': str}. item_index: 0-based item index. Default: 0.

Add a take to a timeline item. Args: media_pool_item_id: Unique ID of the MediaPoolItem to use as take. start_frame: Start frame. Default: 0. end_frame: End frame. Default: 0. item_index: 0-based item index. Default: 0.

Get takes info for a timeline item. Args: item_index: 0-based item index. Default: 0.

Select a take by index. Args: take_index: 1-based take index. item_index: 0-based item index. Default: 0.

Delete a take by index. Args: take_index: 1-based take index. item_index: 0-based item index. Default: 0.

Finalize the selected take. Args: item_index: 0-based item index. Default: 0.

Copy grades from one timeline item to others. Args: target_item_indices: List of 0-based indices of target items. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1. source_item_index: 0-based source item index. Default: 0.

Enable or disable a timeline item. Args: enabled: True to enable, False to disable. item_index: 0-based item index. Default: 0.

Update sidecar file for a timeline item. Args: item_index: 0-based item index. Default: 0.

Load a burn-in preset for a timeline item. Args: preset_name: Burn-in preset name. item_index: 0-based item index. Default: 0.

Create a Magic Mask on a timeline item. Args: mode: 'Forward' or 'Backward'. Default: 'Forward'. item_index: 0-based item index. Default: 0.

Regenerate Magic Mask on a timeline item. Args: item_index: 0-based item index. Default: 0.

Stabilize a timeline item. Args: item_index: 0-based item index. Default: 0.

Apply Smart Reframe to a timeline item. Args: item_index: 0-based item index. Default: 0.

Get voice isolation state for a timeline item. Args: item_index: 0-based item index. Default: 0. track_type: 'audio' or 'video'. Default: 'audio'. track_index: 1-based track index. Default: 1.

Set voice isolation state for a timeline item. Args: state: Dict with isEnabled (bool) and amount (0-100). item_index: 0-based item index. Default: 0. track_type: 'audio' or 'video'. Default: 'audio'. track_index: 1-based track index. Default: 1.

Reset node colors for all nodes in the active clip version. Args: item_index: 0-based item index. Default: 0. track_type: 'video' or 'audio'. Default: 'video'. track_index: 1-based track index. Default: 1.

Get the color node graph for a timeline item. Args: item_index: 0-based item index. Default: 0.

Get the color group for a timeline item. Args: item_index: 0-based item index. Default: 0.

Assign a timeline item to a color group. Args: group_name: Name of the color group. item_index: 0-based item index. Default: 0.

Remove a timeline item from its color group. Args: item_index: 0-based item index. Default: 0.

Export LUT from a timeline item. Args: export_type: LUT type ('EXPORT_LUT_17PTCUBE', 'EXPORT_LUT_33PTCUBE', 'EXPORT_LUT_65PTCUBE', 'EXPORT_LUT_PANASONICVLUT'). path: Output file path. item_index: 0-based item index. Default: 0.

Get items linked to a timeline item. Args: item_index: 0-based item index. Default: 0.

Get the track type and index for a timeline item. Args: item_index: 0-based item index. Default: 0.

Get source audio channel mapping for a timeline item. Args: item_index: 0-based item index. Default: 0.

Get stereo convergence values for a timeline item. Args: item_index: 0-based item index. Default: 0.

Get stereo floating window parameters. Args: eye: 'left' or 'right'. Default: 'left'. item_index: 0-based item index. Default: 0.

Get the MediaPoolItem associated with a timeline item. Args: item_index: 0-based item index. Default: 0.

Get cache status for a timeline item. Args: item_index: 0-based item index. Default: 0.

Enable/disable color output cache for a timeline item. Args: enabled: True to enable, False to disable. item_index: 0-based item index. Default: 0.

Enable/disable Fusion output cache for a timeline item. Args: enabled: True to enable, False to disable. item_index: 0-based item index. Default: 0.

Get a Fusion composition from a timeline item by name. Args: comp_name: Name of the Fusion composition to retrieve. track_type: Track type ('video', 'audio', 'subtitle'). track_index: Track index (1-based). item_index: Item index on the track (0-based).

Link a proxy media file to a clip. Args: clip_name: Name of the clip to link proxy to proxy_file_path: Path to the proxy media file

Unlink proxy media from a clip. Args: clip_name: Name of the clip to unlink proxy from

Replace a clip with another media file. Args: clip_name: Name of the clip to be replaced replacement_path: Path to the replacement media file

Transcribe audio for a clip. The transcription language follows the project's Speech Recognition setting (Project Settings > Subtitles), not a per-call argument. Args: clip_name: Name of the clip to transcribe use_speaker_detection: Optional (Resolve 21+). Enable speaker detection. If omitted, the project's setting is used.

Clear audio transcription for a clip. Args: clip_name: Name of the clip to clear transcription from

Get metadata for a Media Pool clip. Args: clip_id: Unique ID of the clip. metadata_type: Specific metadata key, or empty for all metadata.

Set metadata on a Media Pool clip. Args: clip_id: Unique ID of the clip. metadata: Dict of metadata key-value pairs to set.

Get third-party metadata for a clip. Args: clip_id: Unique ID of the clip. metadata_key: Specific key, or empty for all.

Set third-party metadata on a clip. Args: clip_id: Unique ID of the clip. metadata: Dict of metadata key-value pairs.

Add a marker to a Media Pool clip. Args: clip_id: Unique ID of the clip. frame_id: Frame number for the marker. color: Marker color (Blue, Cyan, Green, Yellow, Red, Pink, Purple, Fuchsia, Rose, Lavender, Sky, Mint, Lemon, Sand, Cocoa, Cream). name: Marker name. note: Marker note. Default: empty. duration: Marker duration in frames. Default: 1. custom_data: Custom data string. Default: empty.

Get all markers on a Media Pool clip. Args: clip_id: Unique ID of the clip.

Get a marker by its custom data string. Args: clip_id: Unique ID of the clip. custom_data: Custom data string to search for.

Update the custom data of a clip marker. Args: clip_id: Unique ID of the clip. frame_id: Frame number of the marker. custom_data: New custom data string.

Get the custom data of a clip marker at a specific frame. Args: clip_id: Unique ID of the clip. frame_id: Frame number of the marker.

Delete all markers of a specific color on a clip. Args: clip_id: Unique ID of the clip. color: Color of markers to delete. Use '' to delete all.

Delete a marker at a specific frame on a clip. Args: clip_id: Unique ID of the clip. frame_id: Frame number of the marker to delete.

Delete a marker by its custom data string. Args: clip_id: Unique ID of the clip. custom_data: Custom data string of the marker to delete.

Add a flag to a Media Pool clip. Args: clip_id: Unique ID of the clip. color: Flag color (Blue, Cyan, Green, Yellow, Red, Pink, Purple, Fuchsia, Rose, Lavender, Sky, Mint, Lemon, Sand, Cocoa, Cream).

Get list of flags on a clip. Args: clip_id: Unique ID of the clip.

Clear flags on a clip. Args: clip_id: Unique ID of the clip. color: Specific color to clear, or empty for all colors.

Get the clip color of a Media Pool item. Args: clip_id: Unique ID of the clip.

Set the clip color of a Media Pool item. Args: clip_id: Unique ID of the clip. color: Color name (Orange, Apricot, Yellow, Lime, Olive, Green, Teal, Navy, Blue, Purple, Violet, Pink, Tan, Beige, Brown, Chocolate).

Clear the clip color of a Media Pool item. Args: clip_id: Unique ID of the clip.

Set a property on a Media Pool clip. Args: clip_id: Unique ID of the clip. property_name: Property name (e.g. 'Clip Name', 'Comments', 'Description'). property_value: Value to set.

Get a property of a Media Pool clip. Args: clip_id: Unique ID of the clip. property_name: Property name, or empty for all properties.

Rename a Media Pool clip. Args: clip_id: Unique ID of the clip. new_name: New clip name.

Link proxy media to a clip. Args: clip_id: Unique ID of the clip. proxy_path: Absolute path to the proxy media file.

Link full resolution media to a proxy clip. Args: clip_id: Unique ID of the clip. full_res_media_path: Absolute path to the full resolution media file.

Unlink proxy media from a clip. Args: clip_id: Unique ID of the clip.

Replace a clip with a new media file. Args: clip_id: Unique ID of the clip to replace. new_file_path: Absolute path to the new media file.

Replace a clip's underlying media while preserving subclip extents. Args: clip_id: Unique ID of the clip to replace. file_path: Absolute path to the replacement media file.

Monitor a growing media file for the given Media Pool clip.

Find a clip by name and return its unique ID. Args: clip_name: Name of the clip to find.

Transcribe audio for a specific clip. Args: clip_id: Unique ID of the clip.

Clear transcription for a specific clip. Args: clip_id: Unique ID of the clip.

Get audio mapping for a clip. Args: clip_id: Unique ID of the clip.

Get mark in/out points for a clip. Args: clip_id: Unique ID of the clip.

Set mark in/out points for a clip. Args: clip_id: Unique ID of the clip. mark_in: Mark in frame number. mark_out: Mark out frame number.

Clear mark in/out points for a clip. Args: clip_id: Unique ID of the clip.

Classify a clip's audio into categories and subcategories (Resolve 21+). Args: clip_id: Unique ID of the clip.

Clear a clip's audio classification (Resolve 21+). Args: clip_id: Unique ID of the clip.

Run IntelliSearch analysis on a clip (Resolve 21+, requires AI IntelliSearch Extra). Args: clip_id: Unique ID of the clip. identify_faces: Whether to identify faces. is_better_mode: Use Better mode (vs Faster).

Run Slate analysis on a clip (Resolve 21+, requires AI Slate ID Extra). Args: clip_id: Unique ID of the clip. marker_color: Marker color for detected slates (Blue, Cyan, Green, Yellow, Red, Pink, Purple, Fuchsia, Rose, Lavender, Sky, Mint, Lemon, Sand, Cocoa, Cream).

Render a motion-deblurred copy of a clip (Resolve 21+). Creates a NEW media file; the source clip is not modified. Args: clip_id: Unique ID of the clip. deblur_option: Settings dict (FileName, Format, Codec, EncodingProfile, UseExtremeMode, UseMarkInMarkOut, RenderAtSourceRes, UseMoreGpuMemory).

Resolve a Media Pool timeline entry to its timeline summary (Resolve 21.0.4+). Calls MediaPoolItem.GetTimeline(). Returns is_timeline=false for ordinary clips — that is an answer, not a failure. Args: clip_id: Unique ID of the Media Pool item.