The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the Python subprocess module and PowerShell scripts to perform repository management and development tasks, such as creating git worktrees, calculating diffs, and executing project setup tools like npm, cargo, and pip.
[COMMAND_EXECUTION]: The recursive-router subskill constructs and executes shell commands for external CLI tools (e.g., Codex, Kimi, OpenCode) to delegate bounded implementation and audit tasks based on user-editable configuration templates.
[PROMPT_INJECTION]: The recursive-training subskill implements an experiential memory system that parses agent-authored markdown artifacts from completed runs to extract reasoning patterns. These patterns are stored in the repository and re-injected as context in future agent sessions, creating a surface for indirect prompt injection if malicious content is processed into the memory plane.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for extending its capabilities by discovering and adding additional AI agent skills via the npx skills command.
[COMMAND_EXECUTION]: The skill implements a cryptographic locking mechanism using recursive-lock.py and verify-locks.py, which calculate SHA-256 hashes of workflow artifacts to ensure process integrity and detect tampering.

recursive-mode