This MCP server provides a sandboxed workspace for agentic applications with tools for reading and writing files, editing files, running shell commands, and listing directory contents. All operations are scoped to a single workspace root directory with strict path validation. Built on FastMCP, it supports both Cursor via MCP and direct Python library integration, keeping LLM and API keys in user-owned applications.
Cognium trust score
64%
Tier
Scanned
Composite of vulnerability cleanliness, spec conformance, provenance, stability, and usage signals — scanned and weighted by Cognium. Human and agent signals are tracked separately.
Last scanned 2026-09-01.
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.