update-claude-learnings
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes session data to update
CLAUDE.md, a file that influences the agent's behavior in future sessions. - Ingestion points: The skill reads the current session history and the project's existing
CLAUDE.mdfile (SKILL.md). - Boundary markers: No explicit delimiters are used to separate user-provided session data from the instruction extraction process.
- Capability inventory: The skill uses
EditandWritetools to modify project files (SKILL.md). - Sanitization: The skill mitigates risks through a mandatory human-in-the-loop confirmation step using the
AskUserQuestiontool before any writes occur. - [COMMAND_EXECUTION]: The skill uses standard file system tools (
Read,Edit,Write,Glob,Grep) to manage documentation. No execution of arbitrary shell commands or external binaries was detected.
Audit Metadata