The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references installation scripts for well-known and trusted development tools including Astral (uv), Bun, and Homebrew from their official domains. These references are part of the environment setup documentation for users.
[COMMAND_EXECUTION]: The agent is instructed to use package managers such as bun and uv to manage dependencies and execute scripts within the target skill's directory.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it reads and adapts instructions from external, potentially untrusted skills. * Ingestion points: The agent reads SKILL.md, AGENTS.md, and source code from a user-provided skill_path. * Boundary markers: The protocol lacks explicit markers or 'ignore' instructions to isolate untrusted data from the agent's logic. * Capability inventory: The agent can perform file system operations and execute shell commands (bun, uv). * Sanitization: There is no evidence of sanitization or validation of the content found within the files being adapted.

tentacle-skills-washing