felo-twitter-writer
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external tweet content via the felo-x-search tool and incorporates it into prompts for style analysis. This is a primary function of the skill.\n
- Ingestion points: The agent fetches tweets using run_x_search.mjs as described in SKILL.md.\n
- Boundary markers: The content is preceded by a "Tweets:" header but lacks strict sandboxing delimiters.\n
- Capability inventory: The skill interacts with the felo-superAgent script and manages documents via felo-livedoc.\n
- Sanitization: No explicit sanitization of fetched tweet text is performed before prompt interpolation.\n- [DYNAMIC_EXECUTION]: The skill instructions utilize node -e to execute inline JavaScript for parsing JSON tool outputs and constructing parameters for the felo-superAgent script. This is used for benign data formatting tasks.
Audit Metadata