session-flow
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user messages and tool results to summarize the session's history.\n
- Ingestion points: Reconstructs the conversation flow from "user messages and tool results" found in SKILL.md.\n
- Boundary markers: The skill does not define specific delimiters for cited user content, although it uses specific internal markers to identify its own tasks.\n
- Capability inventory: The skill utilizes task management tools including TaskList, TaskGet, TaskCreate, and TaskUpdate.\n
- Sanitization: No explicit sanitization or filtering of user-provided quotes is specified before they are incorporated into the summary map.
Audit Metadata