slack
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents a vulnerability surface for indirect prompt injection through its Slack interaction capabilities.
- Ingestion points: External and untrusted data enters the agent context via the
readMessagesaction defined in SKILL.md. - Boundary markers: There are no boundary markers or delimiters defined to distinguish Slack message content from the agent's instructions.
- Capability inventory: The skill provides the agent with the ability to send, edit, and delete messages, as well as retrieve member information, which could be abused if the agent is manipulated by message content.
- Sanitization: No sanitization or validation of the ingested Slack message content is mentioned.
Audit Metadata