The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git and the GitHub CLI (gh) to retrieve repository metadata, PR diffs, and issue details. These are standard operations for a development-oriented skill and align with the allowed-tools configuration.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, specifically PR bodies, issue descriptions, and review thread comments, to generate local documentation.
Ingestion points: Data is fetched via gh pr view, gh issue view, and the GitHub GraphQL API (SKILL.md Steps 1, 2, and 5).
Boundary markers: The skill does not explicitly define delimiters or instructions to ignore embedded commands within the fetched content when summarizing it.
Capability inventory: The skill possesses Write and Edit permissions to modify files in the .claude/skills/branch-context/ directory and executes a local script append-pr-decision.sh.
Sanitization: No explicit sanitization or filtering of the fetched content is described; the skill relies on the agent to "classify" and "summarize" the data before writing it to markdown files.

adopt-pr