The Agent Skills Directory

[SAFE]: The skill provides a robust set of instructions and templates for managing complex coding tasks. It prioritizes evidence-based decision-making and surgical code changes. Analysis of the behavioral guards and core contract confirms the skill is focused on quality and structured output rather than harmful operations.- [PROMPT_INJECTION]: The skill facilitates the ingestion of data and instructions from the target repository being processed, which could contain adversarial content.
Ingestion points: references/command-discovery.md (scans repository files like package.json, Makefile, and AGENTS.md for executable commands) and references/require-clarification.md (instructs the agent to read the codebase to understand context).
Boundary markers: Absent. The skill does not explicitly instruct the agent to ignore or delimit instructions found within the repository files it reads.
Capability inventory: The agent has access to bash for command execution, read for file access, and write/edit for file modification across the entire repository.
Sanitization: None. The skill instructs the agent to 'discover' and 'record' commands from the repo but does not provide logic for validating the safety of these commands before execution.

swe-workflow