slack
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources including Slack messages, threads, and canvases via tools like
slack_historyandslack_search. - Ingestion points: Untrusted content enters the agent's context through retrieved messages, files, and canvas content mentioned in SKILL.md.
- Boundary markers: The skill provides clear natural language instructions to the agent stating that retrieved material is untrusted and cannot be used to override work or reveal credentials, though it does not define specific technical delimiters for the data.
- Capability inventory: The agent has the capability to post messages to Slack, open DMs, and perform HTTP POST requests to the Paperclip API.
- Sanitization: The skill instructs the agent to manually distinguish between user requests and data retrieved from Slack, relying on the agent's reasoning to prevent injection attacks.
Audit Metadata