The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected in the skill's instructions or resource files. The skill provides a structured workflow for interacting with official developer tools and services.
[SAFE]: The skill manages potential indirect prompt injection surfaces by validating the source of external data:
Ingestion points: The skill retrieves data from pull request comments via CLI tools (gh, glab, bb, az) in SKILL.md and resources/providers.md.
Boundary markers: The instructions include logic to filter comments and only process those authored by recognized Qodo bot accounts (e.g., pr-agent-pro, qodo-merge[bot]).
Capability inventory: Uses standard Edit and Write tools for local file modifications and Git for version control.
Sanitization: Risk is mitigated through an interactive human-in-the-loop requirement, where users must explicitly approve every code change before it is applied.
[SAFE]: Encourages secure credential management by instructing users to authenticate through official provider CLI tools rather than hardcoding secrets.

qodo-pr-resolver