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 jv-vogler/skills --skill local-reviewgit clone --depth 1 https://github.com/jv-vogler/skillsWrote 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/jv-vogler/skills/local-review)<a href="https://agentmods.dev/skills/jv-vogler/skills/local-review"><img src="https://agentmods.dev/badge/skills/jv-vogler/skills/local-review/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/jv-vogler/skills/local-review"><img src="https://agentmods.dev/badge/skills/jv-vogler/skills/local-review.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 | $0.00078 | $0.01861 |
| Opus 5 | $0.00039 | $0.00931 |
| Sonnet 5 | $0.00016 | $0.00372 |
| Haiku 4.5 | $0.00008 | $0.00186 |
Grade A, and why
local-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 11d 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 — 175 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Local PR review
Review a PR against what its ticket asked for, verify every finding by running code, and write it all to a local scratchfile — never to GitHub.
Flow
- Target — first that applies: given PR number (
gh pr view/diff <N>); current branch's open PR (gh pr view); local diff vs base (git diff <base>...HEAD). Say which you used. - Spec — find a ticket reference anywhere: branch name, PR title, PR description. Teams differ; there is no fixed pattern. Fetch the ticket via the Jira/Atlassian MCP tools. MCP missing or ticket unreachable → the PR description becomes the spec. Neither exists → infer intent from the code and say so in the overview. Beyond ticket-hunting, the PR description is not review input — judge the changes, not the pitch.
- Big picture — establish what the project does and what the touched area is for: README, CLAUDE.md, directory layout, the modules around the change. A diff can be locally correct and still wrong for the system.
- What changed — describe the change at business level: "fixes the duplicate image on the product page", never "renamed a to b, added an if". No code in the overview.
- Scope check — ticket vs changes: matches / misses pieces / does unrelated extras. Fundamental mismatch (solves a different problem) → stop and ask whether to review anyway. Anything less → record it in the overview's verdict line, keep going.
- Find problems — correctness (edge cases, wrong data, unhandled errors, races) and design (should this exist in this shape, consistency with sibling code, silent breaking changes). Ground every candidate in code you actually read: open the definitions the diff calls, check the data model, compare with siblings.
- Verify — every finding gets tested before it's written up, as you go or in one batch at the end, whichever is cheaper. Prefer a failing test that reproduces the bug — include it in the comment; it's the most useful artifact you can hand the author. When a test can't capture it, run the app and interact with it. Testing needs the PR's code: work in place if already on the branch with a clean tree, otherwise a temporary git worktree. Delete temp tests and worktrees after. Genuinely impractical to test (network, third parties)? High-confidence inference is acceptable — hedge honestly in the comment and mark the proof line. Discard whatever fails verification: a killed false positive is the system working.
- Write — fill the template below.
- Humanize — run the prose through the
writing:humanizeskill. No fake-personal voice in either direction ("I really like this PR…"): genuine strengths go under Quality points as factual bullets; the user writes their own compliments from them.
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.
- 11d ago First seen · 175 lines · 78 tokens per session scan A 32ea6575d6bf
local-review is a skill published in the GitHub repository jv-vogler/skills (1 stars, last pushed 16d ago), licensed MIT. It adds 78 tokens to every session and 1,861 once invoked, about $0.0004 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 2026-08-31.
Other skills, from other repositories
pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
phx-pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
pr-threads-address
Address PR review feedback by systematically working through every unresolved PR review thread on the current branch's PR - analyze each comment, make the requested code changes (with tests where useful), commit, and optionally reply and resolve.
han-release
Cut a Han release: update CHANGELOG.md with the changes since the last release, bump and tag every plugin that changed as {plugin-name}--v{version} so a version-constrained dependency can resolve, and publish a GitHub release crediting every merged pull request and closed issue to the people behind it. Use when…
han-update-documentation
Update Han plugin documentation so every skill, agent, guidance doc, index, and cross-reference is current and accurate. On a non-default branch, scopes the pass to entities the branch actually touched. On the default branch, performs a full documentation sweep across the whole plugin. Use when updating, refreshing…
timeline
Group the current branch's commit history into semantic phases along a timeline, and generate two review documents: architecture evolution + code review.