greptile-config
Pass
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a structured workflow for AI agents to analyze repository structures and tech stacks without executing external code or scripts.
- [SAFE]: Includes extensive documentation for security-critical rules, such as preventing command injection in Model Context Protocol (MCP) servers, ensuring IPC security in Tauri applications, and enforcing authentication boundaries in Next.js dashboards.
- [SAFE]: Provides a validation checklist for the agent to ensure generated JSON configurations are syntactically correct and adhere to Greptile's specifications.
- [SAFE]: No unauthorized data access or exfiltration patterns detected. The instructions prioritize using parameterized queries and structured logging, and it identifies sensitive areas like payments or auth for stricter rule calibration.
- [SAFE]: Uses a 'patternRepositories' feature that references external repositories in the standard 'org/repo' format for cross-repository context, which is an intended feature of the target platform and does not involve direct remote code execution by the agent.
Audit Metadata