L’interface est localisée ; les présentations et documents des Skills sont traduits automatiquement par IA dans votre langue. Le code, les commandes, chemins et termes techniques restent en version originale, et le document de l’auteur fait foi.
Explorer les Skills
Développement · SKILL.md

agent-bom-discover

agent-bom-discover

Discover AI agents, MCP servers, and configurations on this machine or environment. Use when: "find agents", "what's configured", "doctor", "what MCP servers", "show me what's installed", "mcp inventory".

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentation de la Skill

GitHub

name: agent-bom-discover description: >- Discover AI agents, MCP servers, and configurations on this machine or environment. Use when: "find agents", "what's configured", "doctor", "what MCP servers", "show me what's installed", "mcp inventory". version: 0.75.10 license: Apache-2.0 compatibility: >- Requires Python 3.11+. Install via pipx or pip. No credentials required. Registry data (427+ MCP servers) is bundled in-package with zero network calls. metadata: author: msaad00 homepage: https://github.com/msaad00/agent-bom source: https://github.com/msaad00/agent-bom pypi: https://pypi.org/project/agent-bom/ scorecard: https://securityscorecards.dev/viewer/?uri=github.com/msaad00/agent-bom tests: 6533 install: pipx: agent-bom pip: agent-bom docker: ghcr.io/msaad00/agent-bom:0.75.10 openclaw: requires: bins: [] env: [] credentials: none credential_policy: "Zero credentials required. Discovery reads only structural config data (server names, commands, args, URLs). Env var values are replaced with REDACTED by sanitize_env_vars() before any processing." optional_env: [] optional_bins: [] emoji: "\U0001F50E" homepage: https://github.com/msaad00/agent-bom source: https://github.com/msaad00/agent-bom license: Apache-2.0 os: - darwin - linux - windows credential_handling: "Env var values are NEVER extracted from config files. sanitize_env_vars() replaces all env values with REDACTED BEFORE any config data is processed or stored. Source: https://github.com/msaad00/agent-bom/blob/main/src/agent_bom/security.py#L159" data_flow: "Purely local. Reads MCP client config files across 22+ AI tools. Only structural data (server names, commands, URLs) is extracted. Env var values are redacted before processing. No data leaves the machine." file_reads: # Claude Desktop - "/Library/Application Support/Claude/claude_desktop_config.json" - "/.config/Claude/claude_desktop_config.json" # Claude Code - "/.claude/settings.json" - "/.claude.json" # Cursor - "/.cursor/mcp.json" - "/Library/Application Support/Cursor/User/globalStorage/cursor.mcp/mcp.json" # Windsurf - "/.windsurf/mcp.json" # Cline - "/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json" # VS Code Copilot - "/Library/Application Support/Code/User/mcp.json" # Codex CLI - "/.codex/config.toml" # Gemini CLI - "/.gemini/settings.json" # Goose - "/.config/goose/config.yaml" # Continue - "/.continue/config.json" # Zed - "/.config/zed/settings.json" # Roo Code - "/Library/Application Support/Code/User/globalStorage/rooveterinaryinc.roo-cline/settings/cline_mcp_settings.json" # Amazon Q - "/Library/Application Support/Code/User/globalStorage/amazonwebservices.amazon-q-vscode/mcp.json" # JetBrains AI - "/Library/Application Support/JetBrains/*/mcp.json" - "/.config/github-copilot/intellij/mcp.json" # Junie - "/.junie/mcp/mcp.json" # GitHub Copilot CLI - "/.copilot/mcp-config.json" # Tabnine - "/.tabnine/mcp_servers.json" # Cortex Code (Snowflake) - "/.snowflake/cortex/mcp.json" - "/.snowflake/cortex/settings.json" # Snowflake CLI - "/.snowflake/connections.toml" - "~/.snowflake/config.toml" # Project-level configs - ".mcp.json" - ".vscode/mcp.json" - ".cursor/mcp.json" file_writes: [] network_endpoints: [] telemetry: false persistence: false privilege_escalation: false always: false autonomous_invocation: restricted

agent-bom-discover — AI Agent & MCP Server Discovery

Discovers AI agents, MCP servers, and their configurations across 22+ AI tools on this machine. Shows what's installed, configured, and registered in the bundled 427+ MCP server registry.

Install

pipx install agent-bom
agent-bom agents             # discover all AI agents and MCP servers
agent-bom doctor             # check prerequisites and configuration health
agent-bom mcp inventory      # list all discovered MCP servers
agent-bom where              # show all discovery paths

When to Use

  • "find agents" / "what agents are configured"
  • "what MCP servers do I have"
  • "what's configured on this machine"
  • "mcp inventory"
  • "doctor" / "check my setup"
  • "show me what's installed"

Commands

# Discover all AI agents and MCP servers
agent-bom agents

# Check configuration health and prerequisites
agent-bom doctor

# List MCP server inventory
agent-bom mcp inventory

# Show all discovery paths
agent-bom where

Examples

# Discover all agents
agents()

# Run health check
doctor()

# List MCP inventory
inventory()

Example output:

Discovered 3 MCP clients:
  Claude Desktop  — 4 servers configured
  Cursor          — 2 servers configured
  VS Code         — 1 server configured

Servers: filesystem, brave-search, github, slack, postgres, linear, notion
Registry: 5/7 servers found in registry (427+ entries)

Guardrails

  • Discovery is read-only — no files are modified, no packages installed.
  • Env var values in config files are always replaced with ***REDACTED*** before processing.
  • Only structural data (server names, commands, URLs) is extracted.
  • Confirm with user before scanning paths outside the home directory.
Skills enregistrées