Independently assess remaining review findings to triage genuine issues from nitpicks
From mach10npx claudepluginhub leanandmean/mach10 --plugin mach10<pr-number> [--review-comment <id>]opusYou are performing an independent assessment of the most recent PR review findings. The goal is to determine which remaining issues are genuine problems versus nitpicks or false positives, so the PR can proceed to merge with confidence.
User input: $ARGUMENTS
The user's input contains:
--review-comment <id> flag with a numeric comment ID (optional)Example inputs:
108108 --review-comment 1234567890Extract the PR number. Parse the --review-comment flag if present (followed by a numeric ID). If the input is ambiguous, ask the user to clarify.
Read the PR title and description:
gh pr view <pr-number>
If --review-comment was provided: Fetch the specific comment by ID:
gh api repos/:owner/:repo/issues/comments/<review-comment-id>
Extract the body field from the JSON response. This is the review comment content. Note the comment ID for use in Step 6.
If --review-comment was NOT provided (fallback): Fetch all comments as JSON:
gh pr view <pr-number> --json comments
Parse the JSON array and search from the END (most recent first) for the first comment whose body contains the HTML marker <!-- mach10-review -->. If no comment contains the marker, fall back to finding the last comment with the structured review format (Critical/Important/Suggestions sections and model attribution). If a review comment is found, extract its numeric ID from the url field (the number after issuecomment-). Note this ID for use in Step 6.
Fetch the full comment history for context on prior discussion, fixes, and implementation notes:
gh pr view <pr-number> --comments
Understand:
For each finding in the most recent review:
Present your assessment to the user as a clear table or list:
For each finding:
Summarize: how many genuine, how many nitpicks, how many false positives, how many deferred.
Present the assessment to the user, then use AskUserQuestion to ask for approval:
If the user selects "Modify before posting", ask what they want to change, apply the changes, and present the updated assessment for approval again. If the user selects "Skip posting", skip to Step 6.
After the user approves, post a reply comment on the PR documenting:
<!-- mach10-assessment --> as the very first line of the comment body (this invisible HTML marker enables reliable identification in future sessions)When referring to numbered items (findings, suggestions, stages) in the comment body, use plain words like "finding 3" or "suggestion 3" -- not #<number> notation, which GitHub auto-links to issues/PRs.
gh pr comment <pr-number> --body "..."
This comment serves as an audit trail for the human reviewer, demonstrating that each finding was considered rather than ignored.
After posting, retrieve the URL of the assessment comment:
gh pr view <pr-number> --json comments --jq '.comments[-1].url'
Extract the numeric comment ID from the URL (the number after issuecomment-). Note this ID for use in Step 6.
CLI output only (do NOT include in the GitHub comment from Step 5).
First, if an assessment comment was posted and its ID was captured, display the comment IDs for reference:
<review-comment-id from Step 2><assessment-comment-id from Step 5>Then, based on the assessment, recommend next step to the user:
/clear then /mach10:pr-review-fix <pr-number> --review-comment <review-comment-id> --assessment-comment <assessment-comment-id> <issue-numbers>"/clear then /mach10:pr-review-fix <pr-number> --review-comment <review-comment-id> <issue-numbers>"/clear then /mach10:pr-review-fix <pr-number> <issue-numbers>"/mach10:issue-create, then /clear and /mach10:pr-pre-merge <pr-number>"/clear then /mach10:pr-pre-merge <pr-number>"