threads
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Threads interactions, which could potentially contain malicious instructions.
- Ingestion points: Data is retrieved from external Threads users via the list_replies, list_mentions, and open_post tools.
- Boundary markers: No specific delimiters or safety warnings are provided to instruct the agent to ignore instructions embedded within external data.
- Capability inventory: The skill utilizes sensitive tools including publish, reply, and select_channels to interact with the Threads platform.
- Sanitization: The skill focuses on stylistic consistency but does not specify security-based sanitization for the data received from external sources.
Audit Metadata