continuous-learning
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes session transcripts which can contain untrusted data from user input or external web content. 1. Ingestion points: The script evaluate-session.sh identifies session history via the CLAUDE_TRANSCRIPT_PATH environment variable. 2. Boundary markers: The skill does not implement delimiters or instructions to ignore malicious content within the session transcript during analysis. 3. Capability inventory: The agent is tasked with writing new skill files to ~/.claude/skills/learned/ based on its evaluation of the transcript. 4. Sanitization: No sanitization or content validation is performed on the transcript before it is processed for pattern extraction.
- [EXTERNAL_DOWNLOADS]: The skill provides a link to an external documentation resource on X.com.
Audit Metadata