DOC-6771 added skill and slash command to address PR comments

[andy-stark-redis]

This is an improvement on the existing bugbot command in that it reads comments by all contributors or optionally just the ones you pass as parameters to the command (eg, /docs:assess-comments dwdougherty).

Unlike bugbot, it doesn't automatically fix the issues, but gives you a set of suggestions for high-priority fixes. The intention is that you will usually have to use it iteratively and as you do that, you will probably also get new feedback from the bugbot and human reviewers. Once the assessment is finished, you could ask Claude to act on it directly or have it posted to a PR if you are using it as part of a review of someone else's PR.

As usual, all feedback welcome :-)

---

Note

Low Risk
Adds documentation and local Claude command/state only; no runtime product code, auth, or data paths.

Overview
Adds /docs:assess-comments, a report-only Claude slash command that reconciles all PR feedback (humans, Cursor Bugbot, security/history bots, CI, summaries) into one checkpoint report—optionally scoped to a single commenter via $ARGUMENTS.

The workflow pulls inline, issue, and review comments plus GraphQL review thread resolution, clusters findings by code/theme, splits open vs resolved (with fix-quality and “resolved ≠ fixed” checks), deep-verifies only the top 3–5 open clusters per run, and labels patterns (agreement, contradiction, ping-pong, subsystem churn, etc.). It cross-checks human approvals against open bot findings, recommends actions (including when to hand off to /docs:bugbot vs fix manually), and may suggest optional /codex:review—no PR edits except updating .claude/state/assess-comments.coverage.md, an evidence-gated ledger of which heuristics have fired on real PRs.

^{Reviewed by Cursor Bugbot for commit d788bce. Bugbot is set up for automated code reviews on this repo. Configure here.}

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[github-actions]

DOC-6771

[jit-ci]

🛡️ Jit Security Scan Results

✅ No security findings were detected in this PR

---

^{Security scan by Jit}

[dwdougherty]

I pulled this branch and ran /docs:assess-comments . It correctly identified the two Cursor[bugbot] comments. This is pretty cool! I was doing this with a simple prompt, but this is far superior.

I doubt whether all our team members have the GitHub CLI installed (maybe they do), and it's a prerequisite for using this command. Maybe there should be a README describing all the AI tools? Maybe in the repo's README.md file?

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit ba7921d. Configure here.}

[andy-stark-redis]

Thanks @dwdougherty - might come in useful when certain PMs give reviews ;-)