session-info
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns detected. The skill performs intended diagnostic functions for session management.
- [COMMAND_EXECUTION]: Executes a local script via the Bun runtime located within the plugin's own directory structure ($HOME/.claude/plugins/.../get-session-info.ts).
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a local registry file (~/.claude/projects/.../.session-chain-cache.json). This ingestion point is limited to the agent's internal application data, and the registry format documented in registry-format.md uses privacy-preserving hashes (repoHash) instead of full file system paths to identify projects.
Audit Metadata