The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute subprocesses for directory mapping and deep AST analysis using tools like ast-grep to identify exports and service boundaries.- [EXTERNAL_DOWNLOADS]: The skill uses the GitHub CLI (gh) to retrieve file contents and manifest data from remote repositories for version detection and structural analysis.- [PROMPT_INJECTION]: The skill analyzes untrusted repository content, creating an indirect prompt injection surface.
Ingestion points: Project manifests and source code files are read and processed in step-02 and step-04 to identify functional units.
Boundary markers: The instructions do not specify the use of delimiters to isolate the analyzed source code content from the agent's instructions.
Capability inventory: The skill has the capability to execute subprocesses (step-02, step-04) and write files to the local environment (step-06).
Sanitization: In standard operation, a mandatory confirmation gate in step-05 requires the user to review and approve all recommendations before any output is generated; however, this gate is automatically bypassed when headless_mode is enabled.

skf-analyze-source