cc-skill-continuous-learning

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes session transcripts, which contain untrusted user input, to determine if a session is long enough for learning extraction.
  • Ingestion points: The evaluate-session.sh script reads the file specified by the CLAUDE_TRANSCRIPT_PATH environment variable.
  • Boundary markers: No specific boundary markers or 'ignore' instructions are used for the transcript content.
  • Capability inventory: The script performs local file system operations (mkdir) and uses text processing tools (grep, jq, sed).
  • Sanitization: The transcript content is not sanitized, but it is only used for line counting and is not passed to an evaluator or shell command.
  • [COMMAND_EXECUTION]: The evaluate-session.sh script executes standard system commands including mkdir, grep, jq, and sed to process its configuration and analyze session metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 02:32 PM
Security Audit — agent-trust-hub — cc-skill-continuous-learning