The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes git, gh (GitHub CLI), and cargo xtask (a vendor-specific project tool) to manage pull request metadata and contents. These operations are restricted to the local repository and the configured GitHub remote.
[PROMPT_INJECTION]: The skill processes untrusted input from pull request titles, bodies, and git logs which could contain malicious instructions intended to influence the agent's behavior during reconciliation.
Ingestion points: Data enters the environment via gh pr view, gh pr diff, and git log commands as specified in SKILL.md.
Boundary markers: No explicit boundary markers or delimiters are defined for the ingested PR content, though the skill logic segregates prose into specific sections.
Capability inventory: The skill has the ability to write to the local file system (/tmp) and execute shell commands (gh, git, cargo) throughout its process.
Sanitization: The skill explicitly mandates the use of gh pr edit --body-file to prevent shell injection or formatting errors that could occur if PR content containing special characters (like $ or backticks) were passed directly as command-line arguments.

jackin-refresh-pr