Universal MCP server enabling any LLM or AI agent to utilize expert skills from your local filesystem. Skills are structured prompts written once and reused across Claude Code, Claude Desktop, Cline, Continue.dev, or any MCP-compatible client. Features lazy loading (only ~50 tokens per skill initially), hot reload for instant updates, and multi-directory skill aggregation. Skills use portable SKILL.md format with YAML frontmatter.
Cognium trust score
50%
Tier
Unverified
Composite of vulnerability cleanliness, spec conformance, provenance, stability, and usage signals — scanned and weighted by Cognium. Human and agent signals are tracked separately.
Returns 7 tools: search_skills, get_skill, list_leaderboard, get_trust_breakdown, resolve_composition, plus the ChatGPT-connector search and fetch. Every tool is annotated read-only.
Resolve this skill directly via MCP tools/call get_skill.