slack-openapi-skill
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads message history and replies from Slack channels, which are untrusted external sources. * Ingestion points: Data enters the agent context through
/conversations.historyand/conversations.repliesoperations defined inreferences/slack-web.openapi.json. * Capability inventory: The skill can post messages (chat.postMessage) and add reactions (reactions.add), creating a read-write feedback loop. * Boundary markers: No explicit boundary markers or "ignore instructions" directives are used when processing Slack message content. * Sanitization: No sanitization or validation of the retrieved message text is specified before it is processed by the agent. - [EXTERNAL_DOWNLOADS]: Fetches a curated OpenAPI schema from the vendor's GitHub repository to configure the API interface. * Evidence: Found in
SKILL.mdandreferences/usage-patterns.md. * URL:https://raw.githubusercontent.com/holon-run/uxc/main/skills/slack-openapi-skill/references/slack-web.openapi.json
Audit Metadata