The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary function involves executing shell commands to facilitate prototyping. Evidence: The 'Build' section (SKILL.md) explicitly instructs the agent to use the terminal tool to create directories, change working directories, and run dynamically generated scripts (e.g., python3 main.py or npm install).- [EXTERNAL_DOWNLOADS]: The skill mentions external resources and suggests the installation of development tools. Evidence: The instructions reference the gsd-build/get-shit-done repository on GitHub and mention installing the get-shit-done-cc tool using npx (SKILL.md). These are documented as part of the tool's attribution and optional full-workflow integration.- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection due to its research and build loop. Evidence: 1. Ingestion points: The skill uses web_extract to fetch documentation from external URLs and read_file to inspect local files (SKILL.md). 2. Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions embedded in the external content. 3. Capability inventory: The agent has access to terminal for shell execution, write_file for script creation, and delegate_task for fanning out sub-tasks (SKILL.md). 4. Sanitization: No explicit sanitization or validation logic is specified for data fetched from URLs before it influences the generated 'spike' code.

spike