vp-session-organization
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes conversation history to automatically generate or update session titles, which introduces a potential surface for indirect prompt injection.
- Ingestion points: Conversation context and intended outcomes are processed as described in SKILL.md.
- Boundary markers: Present; the skill instructs the agent to use concrete nouns and respect user-chosen names, limiting the influence of untrusted content.
- Capability inventory: The skill uses session renaming and grouping tools like set_thread_title (Codex) and mcp__ccd_session_mgmt__set_session_title (Claude Code) as detailed in the harness capabilities documentation.
- Sanitization: Present; instructions in SKILL.md and references/session-metadata.md explicitly forbid exposing sensitive details in sidebar metadata and require verification of the intended state through readback.
- [SAFE]: No malicious patterns such as credential theft, remote code execution, or persistence mechanisms were detected. The skill's external references point to official documentation for well-known services (Anthropic, Google, OpenAI) or to the author's own repositories, which are documented as vendor-owned resources.
Audit Metadata