Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add urmzd/dotfiles --skill review-diffgit clone --depth 1 https://github.com/urmzd/dotfilesWrote 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/urmzd/dotfiles/review-diff)<a href="https://agentmods.dev/skills/urmzd/dotfiles/review-diff"><img src="https://agentmods.dev/badge/skills/urmzd/dotfiles/review-diff/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/urmzd/dotfiles/review-diff"><img src="https://agentmods.dev/badge/skills/urmzd/dotfiles/review-diff.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00170 | $0.01371 |
| Opus 5.5 | $0.00068 | $0.00548 |
| Sonnet 5.5 | $0.00034 | $0.00274 |
| Haiku 4.5 | $0.00017 | $0.00137 |
Grade A, and why
review-diff 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 1mo 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 — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Review Diff
Review the working-tree changes against a fixed rubric and emit prioritized, actionable findings. This is the day-to-day review skill that [[assess-quality]] (the quality framework) and [[review-design]] (the design "why" layer) both explicitly hand off to. They set the bar; this skill checks one concrete diff against it.
Why a rubric is mandatory
A review with no rubric drifts into a vibe check and produces a near-empty, low-confidence result (in one observed run, a rubric-less review emitted almost nothing at 0.42 confidence). The five dimensions below are loaded before reading the diff so every change is checked against the same axes. If you cannot load the rubric, say so and stop -- do not emit a confident-looking empty review.
Gather the diff
Review everything in the working tree, not just what is staged:
git status --short # the full picture: staged, unstaged, untracked
git diff --staged # staged changes
git diff # unstaged changes to tracked files
git ls-files --others --exclude-standard # untracked files (read each in full)
If a scope base ref is given (e.g. main), use git diff main...HEAD plus the
working-tree diffs. Read untracked files in full with Read -- they have no "old"
side, so the entire file is the change.
The rubric
Check every change against all five dimensions, in priority order:
| Dimension | Ask | Example findings |
|---|---|---|
| Correctness | Does it do what it claims, including edge cases? | off-by-one, null/None deref, unhandled error path, wrong operator, race, resource leak |
| Security | Could this be abused or leak? | injected input in a query/shell, hardcoded secret, missing authz check, unsafe deserialization, path traversal |
| Tests | Is the new behavior covered? | new branch with no test, asserted-nothing test, deleted test, fixture not updated |
| Scope-creep | Does the diff match its stated intent? | unrelated refactor mixed in, drive-by rename, formatting churn that hides the real change, commented-out code |
| Conventions | Does it match the codebase's existing style? | naming, error-handling pattern, logging style, public-API shape, missing doc on an exported symbol |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 1mo ago First seen · 122 lines · 170 tokens per session scan A 587c2d914f0f
review-diff is a skill published in the GitHub repository urmzd/dotfiles (5 stars, last pushed 2d ago), licensed Apache-2.0. It adds 170 tokens to every session and 1,371 once invoked, about $0.0007 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
pr-review
Intelligent PR code review with dynamic agent allocation based on change types. Analyzes PR changes and generates targeted review tasks.
analyze-pr
Analyze a GitHub/GitLab Pull Request by fetching its diff, reading its web description, and generating a comprehensive document. Use when the user provides a PR link and wants to understand what changes the PR introduces, or asks to review/analyze/summarize a pull request.
review-code
Standalone code review for any scope — PR number, branch, file paths, or current changes. Synthesizes intent, runs 11 parallel specialized agents, and presents actionable findings with a verdict. For PR scope, optionally posts the review to GitHub with inline comments.
review-converge
Autonomously review and fix the current branch or specified scope in iterative rounds until clean, leaving all changes in the working tree. Escalates only genuine design decisions. Ends with a transparent report of every round, every fix, and anything it couldn't resolve.
review-pr-comments-converge
Fetch PR review comments, validate each against the actual code, auto-fix obvious issues, and escalate genuine design decisions. Leaves all changes in the working tree. Use when you want to action reviewer feedback autonomously.
review-prs
Fetch and present a prioritized summary of all PRs awaiting the current user's review across the relevant projects and provide an interactive workflow helping the user phrase and post inline review comments and overall review decisions.