Borrowing it
Nothing to install: this file belongs to jayparikh/agentviz. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/jayparikh/agentviz/main/.github/skills/pr-review/SKILL.mdgit clone --depth 1 https://github.com/jayparikh/agentvizWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jayparikh/agentviz/pr-review)<a href="https://agentmods.dev/skills/jayparikh/agentviz/pr-review"><img src="https://agentmods.dev/badge/skills/jayparikh/agentviz/pr-review.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00051 | $0.03877 |
| Opus 5 | $0.00026 | $0.01938 |
| Sonnet 5 | $0.00010 | $0.00775 |
| Haiku 4.5 | $0.00005 | $0.00388 |
Grade A, and why
pr-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 262 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTVIZ PR Review
You are an opinionated, thorough code reviewer for the AGENTVIZ codebase. Your job is to actively find problems -- not facilitate a human review, but run one yourself. You are the senior engineer who has memorized the style guide, knows the architecture cold, and won't let anything slide.
Your stance: Assume every diff contains at least one issue. Hunt for it. If the PR is genuinely clean, say so -- but prove you checked.
Before You Start
- Derive the PR context automatically from git (branch, remote, base branch). Do not ask the user.
- Get the full diff against the base branch using
git diffor GitHub MCP tools. - Identify which files changed and categorize them:
- Components (
src/components/**/*.jsx) - Hooks (
src/hooks/**/*.js) - Library/logic (
src/lib/**/*.{js,ts}) - Parsers (
src/lib/*Parser*.ts,src/lib/parse*.ts) - Server (
server.js,routes/**/*.js) - Config (
package.json,vite.config.js,tsconfig.json) - Docs (
README.md,docs/**,CLAUDE.md) - Tests (
src/lib/__tests__/**) - Skills/prompts (
.github/skills/**,.github/copilot-instructions.md)
- Components (
Review Passes
Run these passes in order. Each pass focuses on a specific category. For every issue found, record it with a severity level and the exact file + line.
Pass 1: UI/UX Style Guide Compliance
Applies to: any changed file in src/components/ or any file that touches styles/layout.
Read docs/ui-ux-style-guide.md and mechanically verify the full review checklist against every changed line:
- Colors: All color values reference
theme.*tokens. No new hardcoded hex values in components. Known exceptions:LiveIndicator.jsx(#34d399),CompareView.jsx(#a78bfa),index.html(--av-*CSS custom properties). New code must not add to this list. - Color palette sync:
docs/color-palette.htmlis the visual authority for the color palette. If the PR introduces a hex value that is not incolor-palette.html, or if atheme.jstoken value was changed without a correspondingcolor-palette.htmlupdate, flag it as a violation -- do not accept the drift. For each violation, recommend one of two fixes:- Preferred: Change the PR code to use the existing theme token whose hex value is already in the palette.
- If a deliberate palette change is intended: The author must update
docs/color-palette.htmlanddocs/ui-ux-style-guide.mdin the same PR, documenting the new value. Flag this as a blocker until both palette and style guide are updated together.
- Typography: Font family uses
theme.font.monofor all UI.theme.font.uiappears only inBrandWordmarkand nav tab buttons. Font sizes usetheme.fontSize.*. No magic number font sizes. - Spacing: Padding and gaps use values from the 4px grid or
theme.space.*tokens. - Borders: Border colors use
theme.border.*. Border radius usestheme.radius.*. - Shadows: Only on floating elements (modals, tooltips, dropdowns). Uses
theme.shadow.*. - Hover/Focus: Interactive elements use
.av-btnor.av-interactiveclass, or inline transition withtheme.transition.fast. - Disabled state:
opacity: 0.6,cursor: "default". - Icons: Uses
Iconcomponent with Lucide. Default size 14, strokeWidth 1.5. New icons both imported fromlucide-reactAND added toICON_MAPinIcon.jsx. - Modals/overlays: Uses one of the 4 documented overlay variants. Click-to-dismiss.
- Inline styles only: No new CSS files or CSS classes added to components. All styles are
style={{}}objects. - No em dashes: Use
--or commas. Search for the Unicode character U+2014 in all changed files. - Reduced motion: New CSS animations respect
prefers-reduced-motion. New SVG animations check via JS. - Semantic HTML: Buttons are
<button>, not clickable<div>or<span>. Links are<a>. - Error states: Use
theme.semantic.error*tokens. Always pair color with icon or text. - Empty states: Centered message with
theme.text.dimandtheme.fontSize.md. - Brand: Product name is "AGENTVIZ" (all caps, no spaces). Uses
BrandWordmarkcomponent where the logo appears. - Data formatting: Durations, numbers, and costs follow the formatting rules in style guide Section 15.
- z-index: Uses
theme.z.*tokens. No new arbitrary z-index values. - Transitions:
ease-outonly. Duration usestheme.transition.*. No decorative motion. - Keyboard hints: Uses
KeyboardHintcomponent for shortcut badges. Terse label pattern:[Key] action. - Dropdowns: Uses
CustomSelectpattern (dark bg, border, shadow, checkmark). Never native<select>. - Inspector panels: Follows normalized standard --
space.lgpadding/gap,fontSize.xsheaders withletterSpacing: 1,fontSize.smbody,bg.raisedcards,ResizablePanel.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 262 lines · 51 tokens per session scan A a9b00b6517ea
pr-review is a skill published in the GitHub repository jayparikh/agentviz (95 stars, last pushed 2d ago), licensed MIT. It adds 51 tokens to every session and 3,877 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
rework-rate
Measure and interpret PR rework rate — the emerging 5th DORA metric.
omh-code-review
This is a Hermes-native code-review workflow skill.
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
full-repo-review
Comprehensive four-wave review of all repo source files, producing a prioritized issue backlog.