The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from untrusted sources within a repository to update documentation.
Ingestion points: The agent reads content from git log outputs, git diff results, package.json, and existing project documentation files (AGENTS.md and related context files).
Boundary markers: The instructions lack explicit delimiters or instructions for the agent to ignore potentially malicious directions embedded within the audited content (e.g., instructions hidden in commit messages or code comments).
Capability inventory: The skill utilizes significant capabilities including executing shell commands (bash), reading files (Read), modifying files (Edit, Write), and deleting files (rm).
Sanitization: There are no specified procedures for escaping or validating external content before it is interpolated into prompts or used to drive file system modifications.
[COMMAND_EXECUTION]: The skill relies on shell commands to inspect the project environment and apply changes.
Executes git commands (log, diff, rev-parse) to determine baseline changes and project structure.
Uses rg (ripgrep) for searching codebase claims.
Employs rm for the destructive removal of documentation files identified as orphans during the progressive-disclosure audit.
Mentions the use of the gh (GitHub CLI) tool for checking the status of branches, pull requests, and issues.

update-agents-md