azure-functions-feedback
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a strong security posture by requiring the redaction of sensitive information, such as API keys, tokens, and customer data, before processing any session findings.
- [SAFE]: A human-in-the-loop verification step is mandatory. The skill must present a preview of any proposed issue or pull request and obtain explicit user consent before utilizing the GitHub CLI or web UI to submit data.
- [SAFE]: The skill interacts with the official Azure/azure-functions-skills repository. This targeted interaction is consistent with the skill's purpose and the author's identity as Azure.
- [SAFE]: While the skill ingests untrusted data from session history and test outputs (potential Indirect Prompt Injection surface), the risks are effectively mitigated by the manual review gate and explicit data sanitization requirements.
- Ingestion points: Conversation history, session files, and test results (SKILL.md Step 1).
- Boundary markers: Absent in prompt construction, but mitigated by Step 3's mandatory user approval.
- Capability inventory: GitHub issue/PR creation and repository file modifications (SKILL.md Steps 4 and 5).
- Sanitization: Specific instructions to redact secrets, tenant data, and connection strings (SKILL.md Step 1).
Audit Metadata