The Agent Skills Directory

[SAFE]: The skill's primary purpose is to explore and document the structure of a project by reading files and metadata. It generates a repository map and an evidence catalog within a specific project folder (.deep-planning/). This behavior is transparent and aligns with the stated documentation and planning goals.- [SAFE]: No indicators of data exfiltration were found. The skill identifies files using tools like 'rg' (ripgrep) and 'cat' but does not include instructions to transmit this data to external or untrusted network locations. Reference to a GitHub Gist by the author is provided for source provenance only.- [SAFE]: The skill does not perform any sensitive file access (such as SSH keys or cloud credentials) or privileged system modifications. It specifically instructs the agent not to run mutating commands and to favor read-only operations for mapping purposes.- [SAFE]: No obfuscated code, hidden instructions, or remote script execution patterns were identified in the skill logic or metadata.

repo-map-codex