projects-health-check
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is potentially vulnerable to indirect prompt injection due to its core functionality of processing external data from multiple sources.
- Ingestion points: Untrusted data enters the agent context through scans of Jira tickets, Slack messages, Fireflies transcripts, and Confluence pages (defined in SKILL.md under Step 1a, 1b, 1c, and 1d).
- Boundary markers: The instructions do not define clear delimiters or XML tags to wrap external content, nor do they provide explicit "ignore embedded instructions" warnings for the agent when processing these summaries.
- Capability inventory: The skill uses high-privilege read tools including
slack_read_channel,slack_search_public_and_private,fireflies_get_transcripts, andgetVisibleJiraProjectsto collect data. - Sanitization: There is no mention of sanitizing or escaping the retrieved text from Slack threads or meeting transcripts before the synthesis step (Step 2).
- [DATA_EXFILTRATION]: The skill accesses sensitive organizational data across four platforms; however, it explicitly mandates that output is shown in chat only and never written elsewhere automatically, reducing the risk of unauthorized data movement.
Audit Metadata