The Agent Skills Directory

[COMMAND_EXECUTION]: The skill calls scripts/recursive-init.py or scripts/recursive-init.ps1 to initialize project run folders. This is a functional component of the automated workflow designed for the repository environment.
[SAFE]: The skill demonstrates an ingestion surface for Indirect Prompt Injection as it reads untrusted repository data (documented in SKILL.md as STATE.md, DECISIONS.md, MEMORY.md, and source code) to inform its drafting process. This context-gathering behavior is localized to the repository and serves the primary purpose of the skill, with capabilities limited to invoking a local initialization script.
[SAFE]: No external network requests, obfuscated code, or unauthorized file access patterns were identified. The skill maintains restricted operations by drafting specs in temporary artifacts until explicit user approval is granted.

recursive-spec