The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to use npx autoskills and npx skills to detect and install relevant agent skills. This leads to the download and execution of code from the npm registry at runtime without version constraints or integrity verification for these non-standard packages.
[EXTERNAL_DOWNLOADS]: The instructions direct the agent to reference and potentially incorporate content from several external GitHub repositories, including obra/superpowers, mattpocock/skills, and addyosmani/agent-skills.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and process untrusted data from the workspace to infer project state.
Ingestion points: Reads package.json, pyproject.toml, README.md, existing CI configurations, and directory layouts (SKILL.md, Phase 1).
Boundary markers: The skill advises the agent to state assumptions and seek confirmation, but does not specify the use of clear delimiters or instructions to ignore embedded commands within the processed files.
Capability inventory: The agent has the ability to execute shell commands (git, npx, uv, bun), write or modify project files, and configure CI/CD workflows.
Sanitization: No explicit sanitization or validation of the content retrieved from workspace files is described before it is integrated into the agent's decision-making process.

setup