← Index

io.github.Olum289/agentlock

io.github.Olum289/agentlock·v0.2.0·Security
Quality Score
80
/100

Security gateway for AI agents: policy, approval, and audited execution, no secrets shared.

§00  Trust verdict · v1 advisory ·  method
UNVERIFIEDno verdict on file

Verdict not yet evaluated for this tool. The hybrid eval runs adversarial cases first; coverage rolls out as the corpus expands. Until a verdict is recorded, an agent should treat this tool as not-yet-cleared and fall back to its own checks. Method: hybrid eval, four-state verdict, honest limits.

§01  Install
Remote endpoint

Streamable HTTP / SSE endpoint. Add to any MCP client that supports remote servers.

https://app.agentlock.net/api/mcp
§03  MCP Quality Score  ·  methodology
freshness
25
completeness
10
installability
25
documentation
15
stability
5
§04  Alternatives in Security