The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes command-line operations to perform repository analysis. Specifically, it uses git clone to download source code to /tmp/docs-analysis and npm install -g repomix to install the Repomix utility for packing repository content into an AI-friendly format.
[EXTERNAL_DOWNLOADS]: The skill fetches content from external sources, primarily prioritizing context7.com for standardized llms.txt documentation. It also performs broad web searches and fetches documentation directly from various library and framework domains.
[PROMPT_INJECTION]: The skill has an inherent surface for indirect prompt injection because it ingests and processes untrusted technical documentation from the internet.
Ingestion points: Untrusted data enters the agent context via WebFetch, WebSearch, and git clone operations defined in SKILL.md and WORKFLOWS.md.
Boundary markers: The instructions do not specify explicit delimiters or "ignore instructions" warnings to separate external documentation from the agent's internal logic.
Capability inventory: The skill has the capability to execute shell commands (npm, git, repomix) as described in SKILL.md.
Sanitization: No sanitization or validation of the fetched documentation content is mentioned before it is processed by the LLM agents.

docs-seeker