← Back to search
@metrxbot/mcp-server
Metrx MCP Server — AI agent cost intelligence tools for LLM agents. Track spend, optimize models, manage budgets, detect waste, and prove ROI.
npm
C
74.7 / 100
Versions
0.2.5latestMay 29, 2026
0.2.3Mar 14, 2026
0.2.2Mar 12, 2026
0.2.1Mar 12, 2026
0.2.0Mar 11, 2026
+ show 4 moreshow less
0.1.3Mar 10, 2026
0.1.2Mar 5, 2026
0.1.1Mar 5, 2026
0.1.0Mar 3, 2026
0.2.5latestfirst seen Jul 21, 2026
0.1.1first seen Jun 5, 2026
Tools 0
No tools indexed yet.
Permissions 0
No permissions indexed yet.
Scan Findings 62
medium
Tool 'run_cost_leak_scan' claims readOnly but name implies write operation
info
Tool 'generate_roi_audit' annotations are consistent
info
Tool 'get_budget_status' annotations are consistent
info
Tool 'set_budget' annotations are consistent
info
Tool 'update_budget_mode' annotations are consistent
info
Tool 'get_alerts' annotations are consistent
info
Tool 'acknowledge_alert' annotations are consistent
info
Tool 'get_failure_predictions' annotations are consistent
info
Tool 'create_model_experiment' annotations are consistent
info
Tool 'get_experiment_results' annotations are consistent
info
Tool 'stop_experiment' annotations are consistent
info
Tool 'configure_alert_threshold' annotations are consistent
info
Tool 'get_upgrade_justification' annotations are consistent
info
Tool 'get_optimization_recommendations' annotations are consistent
info
Tool 'apply_optimization' annotations are consistent
info
Tool 'route_model' annotations are consistent
info
Tool 'compare_models' annotations are consistent
info
Tool 'attribute_task' annotations are consistent
info
Tool 'get_task_roi' annotations are consistent
info
Tool 'get_attribution_report' annotations are consistent
info
Tool 'get_cost_summary' annotations are consistent
info
Tool 'list_agents' annotations are consistent
info
Tool 'get_agent_detail' annotations are consistent
medium
OAuth implementation without PKCE
info
Sandbox failed to start for behavioral verification
medium
Vulnerable dependency: @modelcontextprotocol/sdk@1.12.0 (GHSA-345p-7cg4-v4c7)
high
Permission: shell access detected
medium
Vulnerable dependency: @modelcontextprotocol/sdk@1.12.0 (GHSA-8r9q-7v3j-jr4g)
medium
Vulnerable dependency: @modelcontextprotocol/sdk@1.12.0 (GHSA-w48q-cv73-mx4w)
medium
Vulnerable dependency: vitest@1.6.0 (GHSA-5xrq-8626-4rwp)
medium
Vulnerable dependency: vitest@1.6.0 (GHSA-9crc-q9x8-hgqq)
info
package.json metadata
info
Tool: run_cost_leak_scan
info
Tool: generate_roi_audit
info
Tool: get_budget_status
info
Tool: set_budget
info
Tool: update_budget_mode
info
Tool: get_alerts
info
Tool: acknowledge_alert
info
Tool: get_failure_predictions
info
Tool: create_model_experiment
info
Tool: get_experiment_results
medium
Permission: network access detected
info
Tool: stop_experiment
info
Tool: configure_alert_threshold
info
Tool: get_upgrade_justification
info
Tool: get_optimization_recommendations
info
Tool: apply_optimization
info
Tool: route_model
info
Tool: compare_models
info
Tool: attribute_task
info
Tool: get_task_roi
info
Tool: get_attribution_report
info
Tool: get_cost_summary
info
Tool: list_agents
info
Tool: get_agent_detail
info
Transport: streamable-http
info
Required env vars (2)
info
Sandbox failed to start for output poisoning scan
low
Permission: env_vars access detected
info
SBOM generated: 267 components
medium
No build provenance detected (SLSA L0)