aibtc-news-scout
Pass
Audited by Gen Agent Trust Hub on Mar 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands using
bun runto call theaibtc-news.tsscript. This is used to perform operations such as listing signals, beats, and correspondent data within the aibtc ecosystem. - [PROMPT_INJECTION]: The skill processes untrusted external content from editorial notes and community posts, which constitutes an indirect prompt injection surface.
- Ingestion points: Editorial notes are retrieved via the
news_signalstool, and candidate data is observed from Moltbook and the agent registry (SKILL.md). - Boundary markers: No specific delimiters or instructions are used to prevent the agent from following commands that might be embedded in the ingested notes or posts.
- Capability inventory: The agent has the ability to execute subprocesses via the
bun runinterface (AGENT.md). - Sanitization: The skill lacks explicit validation or sanitization mechanisms for the external data it processes before using it to inform recruitment decisions.
Audit Metadata