The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automates the use of project-related shell commands including mix for Elixir compilation and testing, git for version control, and gh (GitHub CLI) for pull request management.
[EXTERNAL_DOWNLOADS]: Dependencies are noted on external tools such as the codex CLI and ralph-loop, which are used to augment code review processes and enable long-running autonomous execution loops.
[PROMPT_INJECTION]: The orchestration logic presents a surface for indirect prompt injection because the agent processes data from web research and the existing codebase to inform automated implementation tasks.
Ingestion points: User-provided feature descriptions, data retrieved by research agents (e.g., web-researcher), and patterns analyzed from the codebase.
Boundary markers: The instructions do not specify the use of delimiters or specific warnings to ignore embedded instructions within processed data.
Capability inventory: The skill possesses extensive capabilities including file system writes, sub-agent spawning, and execution of shell commands.
Sanitization: There is no explicit mention of sanitization or validation logic for data ingested from the project environment or external research.

full