new-learning
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests data from external web sources and local archives, creating an attack surface for indirect prompt injection.
- Ingestion points: Local lesson archives at
~/.claude/learning/lessons/archive/and external web search results for learning resources. - Boundary markers: Absent; the agent is not instructed to isolate or treat external content as strictly untrusted data.
- Capability inventory: The skill has the ability to write to the file system at
~/.claude/learning/backlog.md. - Sanitization: No sanitization or validation of the content retrieved from web searches is performed before processing.
Audit Metadata