← Back to search

@perceptron-ai/mcp-server

GitHub Actions Scanned 13d ago

Perceptron MCP server for high-accuracy visual perception powered by fast, efficient vision-language models

npm ↗
A
97 / 100

Versions

0.2.0latest
May 8, 2026
0.1.5
Mar 26, 2026
0.1.2
Mar 24, 2026
0.1.1
Mar 20, 2026
0.1.0
Mar 19, 2026
PermissionsTool SafetyAuthAnnotationsCode QualityStabilitySpecVuln HistoryAuthorTransparencyCommunity

Tools 5

list_models
annotations: verified low

List available Perceptron models and their capabilities.

readOnlyHint true
question
annotations: verified low

Ask a question about an image or video. Accepts a URL or local file path.

model string question string
readOnlyHint true
caption
annotations: verified low

Generate a caption for an image or video. Accepts a URL or local file path.

model string style enum
readOnlyHint true
ocr
annotations: verified low

Extract text from an image using OCR. Accepts a URL or local file path.

mode enum model string prompt string image_url string
readOnlyHint true
detect
annotations: verified low

Detect objects in an image or video. Accepts a URL or local file path.

model string classes array
readOnlyHint true

Permissions 2

network medium
Server uses network capabilities via: fetch()
env_vars low
Server uses env_vars capabilities via: process.env

Scan Findings 19

medium
Permission: network access detected permission_analyzer · 70%
info
Tool 'list_models' annotations are consistent annotation_checker · 80%
info
Tool 'question' annotations are consistent annotation_checker · 80%
info
Tool 'caption' annotations are consistent annotation_checker · 80%
info
Tool 'ocr' annotations are consistent annotation_checker · 80%
info
Tool 'detect' annotations are consistent annotation_checker · 80%
info
Sandbox failed to start for behavioral verification behavioral_verifier · 100%
info
package.json metadata manifest_parser · 100%
info
Tool: list_models manifest_parser · 85%
info
Tool: question manifest_parser · 85%
info
Tool: caption manifest_parser · 85%
info
Tool: ocr manifest_parser · 85%
info
Tool: detect manifest_parser · 85%
info
Transport: streamable-http manifest_parser · 80%
info
Required env vars (2) manifest_parser · 80%
info
Sandbox failed to start for output poisoning scan output_poisoning · 100%
low
Permission: env_vars access detected permission_analyzer · 90%
info
SBOM generated: 274 components sbom_generator · 100%
medium
No build provenance detected (SLSA L0) slsa_assessor · 90%