performance-analyzer

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script located at ${CLAUDE_PLUGIN_ROOT}/scripts/connectors/youtube.py to fetch metrics such as views, likes, and comments for campaign videos.
  • Evidence: Found in the "Measured YouTube post-performance" section of SKILL.md.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from external sources like influencer reports and social media comments.
  • Ingestion points: Untrusted data enters the agent context through influencer-supplied reports/screenshots (Step 1) and social media comments used for sentiment analysis (Step 6).
  • Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore embedded instructions within the processed data.
  • Capability inventory: The skill has the capability to write data to the file system (e.g., memory/influencer/performance-analyzer/) and promote facts to a shared memory file (memory/hot-cache.md).
  • Sanitization: There is no evidence of sanitization, validation, or filtering of the external content before processing or summarizing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 10:02 AM
Security Audit — agent-trust-hub — performance-analyzer