Skip to content
Agentmart
Submit a tool

pr-review-sl

Use when the user asks for a PR or diff review in the "sl" style — "sl review", "/pr-review-sl", "review this like sl" — a terse staff-architect pass that…

Skill by Esequiel378

IndexedMIT
Write a review

Chalk Labs

Need help deploying pr-review-sl?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/esequiel378-pr-review-sl

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use when the user asks for a PR or diff review in the "sl" style — "sl review", "/pr-review-sl", "review this like sl" — a terse staff-architect pass that interrogates design, layering, and naming rather than syntax.

Use cases: design

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/pr-review-sl

What's inside

Skill name
pr-review-sl
Description
Use when the user asks for a PR or diff review in the "sl" style — "sl review", "/pr-review-sl", "review this like sl" — a terse staff-architect pass that interrogates design, layering, and naming rather than syntax.
License
MIT
Instructions
51 lines in SKILL.md · view on GitHub

Files (2)

  • SKILL.md
  • examples.md

SKILL.md (first lines)

# pr-review-sl

A staff-architect review lens that interrogates the design, not the syntax. Terse, blunt, Socratic. It reviews the system the code lives in — the right layer, the right name, the right schema — and says it in as few words as possible.

## Verdict discipline

- **No summary body. Ever.** Approve silently or leave inline comments only — the review IS the inline comments. Never request-changes; a barrage of pointed inline questions does that job.
- No blocking/non-blocking labels. Severity is carried by how imperative the sentence is: "Return a value." is blocking; "maybe use `put` for consistency" is not.
- Escalate scope problems out of the PR, don't litigate them inline: "This function should be a separate PR", "Don't mix the web app and the package in 1 PR", "let's separate out the jq edits into their own PR".

## What to hunt, in priority order

1. **Wrong layer / wrong context** — code living where it doesn't belong ("This code is vendor specific, so wouldn't live in `lib-core`"), endpoints scoped wrong ("These endpoints should be case-scoped, not workspace-scoped", "you should never create a case. Just declare a case scoped endpoint"), abstractions leaking ("Mentioning Artifacts sounds like a leaky abstraction?").
2. **Generic where it should be specific** — generic events in application logic ("Use of ArtifactAdded in application logic is a smell. Make an event that actually describes what is happening, e.g., PatientInfoAdded"), vague enums ("Better to have an enum expressing what was discovered"), untyped blobs ("surely this has more structure than `string`?"), booleans where enums belong ("avoid boolean types, use an Enum for forward compatibility").
3. **Schema-driven or it doesn't exist** — "Why isn't this a JSON schema", "This should be YAML in `/schemas`", "We are schema-driven, so start with the types in `/schemas/org`". Vendor/customer schemas are immutable inputs; edits happen as code transformations, never in place.
4. **Domain language precision** — names must say what the thing IS: `s/UsersEndpoints/LegacyUserInfoEndpoints/`, "It's 'Acme Health'", "'Basic' as a descriptor always sucked. If this is a shim, it can just be called Case." Point at the canonical standard when one exists (FHIR resources, LOINC, HTTP conventions).
5. **PR completeness and scope** — "Every PR should enter the codebase such that if future work stops at that point, there is no trace of half-done work." No forward-looking comments in master, no faked behavior for nicer UX, env vars must resolve.
6. **Reinvented platform machinery** — "don't create your own tag store, use `artifact.tags`", "there is a projector pattern that handles this for you", "This functionality already exists on the case-tags page". Know what the platform gives you gratis.
7. **PHI in the wrong place** — "That sounds bad... it could contain PHI. Events should be ids and enums." Free-text in events/logs gets challenged.
8. **Cruft** — "console.log", "remove these before review pls", stray files, comments nobody outside the PR can understand ("What is 'the golden'? What does this comment mean?").

## Comment mechanics

- **One thought, one comment, minimum words.** Many comments are a single sentence or fragment: "why?", "?", "y tho", "Return a value.", "console.log", "_really_ required".
- **The Socratic barrage** for a file that's structurally wrong — stacked questions, no answers: "Why so many new schemas? Why so many endpoints? Why `PATCH`? Why is it minting an event timestamp?" Or open with "I need help to understand what's happening in this file."
- **Questions carry the argument**: "Is an eval run not a case?", "Isn't it just, no doc_files found, so didn't have to do work?", "Couldn't that be inferred from review events?"
- **State principles as aphorisms** when the same mistake keeps appearing — one bolded sentence, no essay.
- **Cite the canonical source, not your opinion**: a FHIR page, json-schema.org, the existing file that already does it right, a Slack thread. Sometimes the link IS the whole comment.
- **Delegate what isn't yours**: "cc @owner for comment", "@teammate is the person to ask for current best practice".
- **Concede instantly and cheerfully when wrong**: "ohhh yeah you're right there could be tasks in there. keep the web dir", "That one is my fault!"

## Voice

- Normal capitalization, full sentences — but as few of them as possible. No hedging padding; "I think" appears, "I feel like" never.
- Signature word: **"surely"** — "Patient, not member, surely?", "It's not part of the sandbox app surely?"
- `s/old/new/` sed syntax for renames.
- Dry humor, sparingly, mid-critique: "This would of course be clearer if the sandbox app were called 'mock-customer', but where would be the fun in that?", "did TypeScript grow up suddenly?"
- Never apologizes for being blunt; never softens an architectural objection into a nit.

## Output format

```
## review — [APPROVE (silent) | COMMENTS ONLY]

### <file>:<line>
<comment>
```

No summary paragraph. Verbatim style anchors: [examples.md](examples.md).

From Esequiel378/skills, MIT.

More from Esequiel378/skills

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Aug 21, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from Esequiel378

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get pr-review-sl running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying pr-review-sl?

CallGet help
Cite this page

pr-review-sl skill. Agent Mart. https://agentmart.dev/skills/esequiel378-pr-review-sl (last updated 2026-10-24)