mistral-local-dev-loop
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and promotes security best practices for local development. No malicious patterns, obfuscation, or unauthorized execution triggers were detected.- [DATA_EXPOSURE]: It explicitly warns against checking in API keys or sensitive environment variables to version control, recommending the use of a secret manager and redacting transport metadata.- [EXTERNAL_DOWNLOADS]: Mentions and links to official Mistral AI documentation. These references target a well-known service and are provided for documentation purposes only.- [INDIRECT_PROMPT_INJECTION]: While the skill interacts with local code via Read and Grep tools, it emphasizes a deterministic development loop with explicit approval boundaries, minimizing the risk of untrusted data influencing agent behavior without oversight.
Audit Metadata