The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands and the GitHub CLI to perform project setup and manage issue dependencies. The script scripts/dependencies.py uses subprocess.run to call the gh API.\n
Evidence: The script scripts/dependencies.py executes gh commands with parameters derived from a JSON plan, using list-based arguments and avoiding shell invocation to prevent command injection.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources such as GitHub issue labels, descriptions, and pull request bodies, which could potentially contain malicious instructions.\n
Ingestion Points: The references/issue-tracker.md file instructs the agent to inventory labels and inspect usage across open and closed issues and PRs.\n
Capability Inventory: The skill can modify project documentation and update GitHub issue relationships via the gh API.\n
Boundary Markers: Core instructions in SKILL.md require the agent to present an approval plan and obtain user consent before applying any mutations.\n
Sanitization: The scripts/dependencies.py script validates repository names using regular expressions and passes structured JSON to the API via standard input.\n- [EXTERNAL_DOWNLOADS]: The skill's documentation provides installation instructions for companion skills and libraries from the author's official GitHub repository.\n
Evidence: README.md includes commands to add companion skills like domain-modeling and github-repo-setup using npx skills add SirDarcanos/agentskills, which are identified as verified vendor resources.

setup-project