The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to utilize a CLI utility named brief to generate a snapshot of the repository's structure and contents.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted data from the user's working directory to inform its recommendations, creating a potential vector for malicious project files to influence agent behavior.
Ingestion points: The skill reads various manifest files (e.g., package.json, pyproject.toml, Cargo.toml), documentation (e.g., README, docs/), and source code entry points.
Boundary markers: The instructions lack specific guidance on using delimiters or "ignore instructions" markers when processing text from the repository files.
Capability inventory: The skill's primary function is to search for and recommend other skills to be loaded into the agent's context. It also leverages the brief CLI tool.
Sanitization: There are no mentioned mechanisms for sanitizing or filtering instructions that might be embedded in README files or code comments meant to trick the researcher into recommending a specific (potentially malicious) skill.

skill-researcher