unslop-review
Rewrites code review comments so they read like a human teammate wrote them. Cuts corporate-AI throat-clearing ("I noticed...", "I was wondering if perhaps...", "It might be worth considering..."). Each comment is direct: location, the issue, a concrete fix. Use when user says "humanize review", "de-slop PR comment", "make this feedback sound human", "review this PR", "code review", "/review", "/unslop-review". Auto-triggers when reviewing PRs.
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-14
- Stars
- ★ 1.6k
- Forks
- 491
01 前言 Front Matter
- name
- unslop-review
- description
- Rewrites code review comments so they read like a human teammate wrote them. Cuts corporate-AI throat-clearing ("I noticed...", "I was wondering if perhaps...", "It might be worth considering..."). Each comment is direct: location, the issue, a concrete fix. Use when user says "humanize review", "de-slop PR comment", "make this feedback sound human", "review this PR", "code review", "/review", "/unslop-review". Auto-triggers when reviewing PRs.
03 SKILL.md 正文
Unslop Review
Purpose
Rewrite or generate PR review comments that sound like a teammate, not a politeness engine. Direct on the issue, concrete on the fix, kind on the human.
Trigger
/unslop-review, /review, "review this PR", "code review", "humanize review", "de-slop this comment", "make this feedback sound human". Auto-trigger when reviewing pull requests.
Format
Default shape: L<line>: <severity prefix> <observation>. <fix>.
Severity prefixes (optional but use them when severity matters):
bug:— code is broken or will breakrisk:— works today, fragile tomorrow (perf, race, missing test)nit:— style, naming, dead code, "while you're here"q:— genuine question, not a hidden complaint
Multi-file: <file>:L<line>: <severity> <observation>. <fix>.
Range: L88-140: ... when the issue spans lines.
Rules
Drop
- Throat-clearing: "I noticed that...", "It seems like...", "It looks like to me..."
- Stacked hedging: "I was wondering if perhaps we might want to potentially..."
- Polite-padding: "I would kindly suggest...", "just a small suggestion..."
- Per-comment praise: "Nice work on this function but...", "Great pattern, however..."
- Restating the diff: "Here on line 42 you have a function called
getUserwhich returns..." - Bare opinion without a fix: "This is bad" with no suggestion
Keep
- Exact line numbers and ranges
- Identifiers in backticks:
findUser,req.body.id - Concrete fix or concrete question
- "Why" only when the fix isn't obvious
Tone
Human, not corporate. "This throws if X" not "It may potentially be worth considering that this could throw under certain conditions." Calibrated uncertainty is fine ("I think", "probably") — performative softening is not.
Auto-clarity (use full prose, not one-liners)
- Security findings (CVE-class, auth, secrets)
- Architecture disagreements that need a real discussion
- Onboarding context for a new contributor
- When the answer is genuinely "this is fine"
In those cases use a short paragraph, then resume terse for the rest.
Examples
Bad → good
-
Bad:
I would kindly suggest that we might want to potentially consider adding a null check here as it could maybe lead to issues in some scenarios. -
Good:
L42: bug: \findUser` returns undefined when no match. Guard before `user.email` or early-return 404.` -
Bad:
Great work on this implementation! However, I think we could potentially enhance readability by considering a refactor of this function. -
Good:
L88-140: nit: this function does validation, I/O, and mapping. Splitting them would make the happy path easier to follow. Happy to pair on a cut if helpful. -
Bad:
I noticed that there's no retry logic here which could be problematic. -
Good:
L23: risk: no retry on 429. Wrap the call in \withBackoff(3)` so we don't drop legitimate requests.` -
Bad:
This implementation leverages a robust caching strategy. -
Good: (delete — empty praise. If the caching is genuinely interesting, explain why specifically.)
Approval
If the change is solid and you have nothing concrete: LGTM on its own line. No boilerplate.
Boundaries
- Comments only. No commits, no
git push, no auto-approve, no linter runs. - Output is paste-ready: one comment per line, or a clearly separated list.
- Severity must be honest. Don't downgrade a
bugto anitto soften the message.
04 同仓库更多 Skills
worktable-review-with-annotations
Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.
worktable-find-and-synthesize-context
Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.
worktable-create-or-update-docs
Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.
worktable-create-or-manage-records
Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.
worktable-create-or-manage-interactive-html
Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.
worktable-collaborate-threads
Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.