The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions for the agent to download and install external code from GitHub repositories via the npx skills add command. This enables the agent to dynamically extend its functionality with third-party logic.- [COMMAND_EXECUTION]: The skill utilizes system commands to perform management tasks. It specifically instructs the agent to use the -g (global) flag for installation and the -y flag to skip interactive prompts, which allows for environment modification without human oversight.- [EXTERNAL_DOWNLOADS]: The skill fetches definitions and code from the skills.sh registry and various GitHub repositories, including the author's own tools and repositories from trusted organizations like Vercel.- [PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection by processing uncontrolled user queries to search for and install external modules. Maliciously crafted search results could influence agent behavior.
Ingestion points: User-provided [query] parameter in SKILL.md.
Boundary markers: Absent; no delimiters are used to isolate external search results.
Capability inventory: File system modification and package installation via npx in SKILL.md.
Sanitization: Absent; no validation or filtering of content from the external registry is described.

find-skills