The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to inspect the local environment and interact with external APIs.\n
Evidence: Uses ls and cat to locate and read the first 80 lines of README.md and package.json for product context (SKILL.md).\n
Evidence: Executes curl to send generated prompts to the Google Gemini API (SKILL.md).\n- [EXTERNAL_DOWNLOADS]: The skill's documentation recommends installation through external package managers and specific registries.\n
Evidence: Recommends installation via npx "@opendirectory.dev/skills" (README.md).\n
Evidence: Suggests installing plugins from opendirectory-marketplace (README.md).\n- [DYNAMIC_EXECUTION]: The skill generates content and scripts at runtime to facilitate its core functionality.\n
Evidence: Generates a JSON payload file in /tmp/ph-launch-request.json and then executes a complex shell command that pipes the output of an API call into a python3 one-liner for JSON parsing (SKILL.md).\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local files which could contain instructions intended to influence the agent's behavior or the API output.\n
Ingestion points: Reads product data from README.md and package.json (SKILL.md).\n
Boundary markers: Employs a structured JSON format for the API request, providing a level of separation between instructions and data.\n
Capability inventory: The skill has the ability to execute shell commands, perform network requests via curl, and write to the filesystem (SKILL.md).\n
Sanitization: No explicit sanitization or filtering is applied to the content extracted from project files before it is included in the generative prompt.

producthunt-launch-kit