The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified in the skill's instructions, configuration, or testing code.
[EXTERNAL_DOWNLOADS]: The ATTRIBUTION.md file contains a reference to the original source repository on GitHub (github.com/mattpocock/skills). This is documented as a neutral reference to the skill's provenance and does not involve automated or unsafe code execution.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze potentially untrusted project data (git history, diffs, test outputs). This inherent ingestion surface is mitigated by the skill's workflow, which generates descriptive reports and requires explicit human approval for any suggested actions, preventing automated execution of instructions found in external data.

project-retro