The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill includes a 'Codebase Analysis Mode' that instructs the agent to ingest untrusted data from a codebase using 'Read' and 'Grep' tools. However, the skill's capabilities are limited to these read-only operations, and it does not perform any sensitive actions (like network exfiltration or file writing) based on the ingested content. The workflow is purely analytical.
[COMMAND_EXECUTION]: The skill utilizes the 'Grep' and 'Read' tools to analyze codebase surfaces. These are standard tools for development-oriented agents and are used here within a well-defined, safe workflow for structural analysis.
[EXTERNAL_DOWNLOADS]: The skill references grounding sources from established organizations (NetworkX, W3C, Barbara Minto) and a repository belonging to the skill's author (jacob-balslev). These references are used for conceptual grounding and documentation, representing safe external references.

semantic-center