From base
Fetch and analyze unresolved PR review comments using a specialized GraphQL script that extracts thread-level data (thread IDs, resolution status, outdated flags, line ranges) not available through basic gh commands. MUST consult this skill whenever the user wants to check review feedback, see what reviewers said, view unresolved threads, understand pending review comments, or prepare to address reviewer requests on a PR. This is the required first step for the fixing-review-comments workflow. Also trigger on "review comments", "レビューコメント", "unresolved threads", "レビュー指摘", "reviewer feedback", "PR feedback", or any request to read, fetch, list, or summarize review discussions on a pull request.
npx claudepluginhub kkhys/claude-code-marketplaceThis skill is limited to using the following tools:
Fetch unresolved review comments from the current PR and produce a fix plan that downstream skills can act on.
Creates isolated Git worktrees for feature branches with prioritized directory selection, gitignore safety checks, auto project setup for Node/Python/Rust/Go, and baseline verification.
Executes implementation plans in current session by dispatching fresh subagents per independent task, with two-stage reviews: spec compliance then code quality.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Fetch unresolved review comments from the current PR and produce a fix plan that downstream skills can act on.
bash "${CLAUDE_SKILL_DIR}/scripts/read-unresolved-pr-comments.sh"
Prerequisite: the current branch must have an open PR (gh pr view must succeed).
The script returns JSON with PR metadata and all unresolved threads. Each thread includes thread_id, path, line, start_line, is_outdated, and the full comment history.
If no unresolved threads exist, report that to the user and stop.
Analyze each thread to understand what the reviewer is actually asking for. Often the surface-level comment ("rename this variable") reflects a deeper concern ("this name is misleading because it suggests X when the value is Y"). Capture both the specific ask and the underlying intent.
Structure the plan as follows:
is_outdated threads explicitly — the referenced code may have already changed, so verify the current state before planning a fixfixing-review-comments may execute them in parallel via subagentsThis skill only reads and plans. It does not modify code, resolve threads, or push changes. The fix plan is a checkpoint for the user or the calling skill to review before any code is touched.