The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/check-fixture.py script utilizes subprocess.run to execute a synthetic test script within a temporary directory for validation purposes. This is a standard testing pattern and does not involve shell execution.
[DYNAMIC_EXECUTION]: The test utility scripts/check-fixture.py generates and runs a Python script from an internal template to simulate the implementation loop. This behavior is contained within the test fixture and does not process untrusted external input.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external code and instructions, which represents an inherent attack surface. Ingestion points: The agent is instructed in SKILL.md to read target instructions, revision agreements, and readiness evidence. Boundary markers: The workflow relies on following authoritative agreements and explicit scope definitions to mitigate risks. Capability inventory: The skill involves file system modifications and the execution of checks via agent tools. Sanitization: The instructions recommend using actual parsers and schemas for configuration validation to ensure safety settings are respected.
[EXTERNAL_DOWNLOADS]: The README.md file contains instructions to install the skill via npx from the author's own repository (SirDarcanos/agentskills), which is a standard procedure for vendor resources.

build-and-check