ShellWard intercepts tool calls at runtime to enforce security guardrails using an 8-layer defense architecture. It provides DLP-style data flow control that blocks outbound data transmission when sensitive information was recently accessed, prompt injection detection with 32 rules covering English and Chinese, dangerous command blocking, PII scanning, and session monitoring. Works as a standalone SDK or MCP server with zero dependencies across Claude Desktop, Cursor, and other MCP-compatible clients.
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.