The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from project revision documents and conversation history to synthesize a final report.
Ingestion points: The skill reads all revision files (.revN.md), review notes (.notes.md), and the session conversation history as described in the Inputs section of SKILL.md.
Capability inventory: The skill performs file system modifications, specifically writing new .final.md files and deleting revision files using git rm, followed by a git commit operation.
Boundary markers: The instructions lack explicit delimiters or specific "ignore embedded instructions" directives to distinguish between documentation content and potential instructions embedded within the processed history.
Sanitization: No specific sanitization, validation, or filtering logic is described for the text extracted from input documents or chat history before it is incorporated into the final output document.
[COMMAND_EXECUTION]: The skill uses shell-based tools to manage files and repository references.
The skill instructs the agent to execute git rm on files identified by the user-provided feature and stem arguments.
The skill uses grep to search the entire repository for inbound links to the removed revision files to facilitate updating references.
These commands are standard for the tool's purpose but represent an interaction with the system shell based on user input.

finalize-doc