Borrowing it
Nothing to install: this file belongs to tomtom-international/tomtom-maps-mcp. 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/tomtom-international/tomtom-maps-mcp/main/.agents/skills/tomtom-mcp-review/SKILL.mdgit clone --depth 1 https://github.com/tomtom-international/tomtom-maps-mcpWrote 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/tomtom-international/tomtom-maps-mcp/tomtom-mcp-review)<a href="https://agentmods.dev/skills/tomtom-international/tomtom-maps-mcp/tomtom-mcp-review"><img src="https://agentmods.dev/badge/skills/tomtom-international/tomtom-maps-mcp/tomtom-mcp-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/tomtom-international/tomtom-maps-mcp/tomtom-mcp-review"><img src="https://agentmods.dev/badge/skills/tomtom-international/tomtom-maps-mcp/tomtom-mcp-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 | $0.00000 | $0.03865 |
| Opus 5.5 | $0.00000 | $0.01546 |
| Sonnet 5.5 | $0.00000 | $0.00773 |
| Haiku 4.5 | $0.00000 | $0.00386 |
Grade C, and why
tomtom-mcp-review scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
git worktree remove --force "$REVIEW_DIR/tree"; rm -rf "$REVIEW_DIR" How it starts
The opening of the file, as written. The whole thing — 234 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A second reviewer for a finished change. The author's agent knows what the change meant to do, so it
reads past what the diff actually does; a fresh subagent sees only the diff, the PR description
and the repository. This skill is the main agent's procedure; the subagent's brief is
docs/reviewer.md.
- Preflight is not this.
tomtom-mcp-preflightruns the CI gates on your working diff before a commit. A review runs no gate, since CI does, and judges the change. - The repository decides. Every comment rests on
CODING_GUIDELINES.md,CONTEXT.md, an ADR,Adding_new_tools.md, or a traced failing scenario — never on taste.
The commands are POSIX shell: Git Bash on Windows. Run each gh command, and fetchPr.mjs, in a
shell call of its own, with no redirect, pipe or $(...). Claude Code's sandbox hides gh's token,
and .claude/settings.json takes these out of it only when the call holds nothing else.
Step 1: Resolve the target
| Invoked with | Mode | Review root |
|---|---|---|
#N, N, or a github.com/tomtom-international/tomtom-maps-mcp/pull/N link |
PR | a worktree at the PR head |
a branch name, --base <ref> optional |
Local | a worktree, or in place when it is the current branch |
| nothing | Local, the current branch, uncommitted changes included | in place |
--all, or --all <path>… |
Local, whole files: every tracked file, or those under the paths | in place |
Add --loop to a Local target on the current branch for loop mode, which Step 4 describes. Refuse
it with a PR, another branch, or uncommitted changes to tracked files.
Refuse a PR link to any other repository. Then set up the inputs in one scratch directory:
REVIEW_DIR=$(mktemp -d "${TMPDIR:-/tmp}/tomtom-mcp-review.XXXXXX")
cp .agents/skills/tomtom-mcp-review/docs/reviewer.md "$REVIEW_DIR/"
In PR mode, fetch the PR with $TARGET, what the user gave: #N, N or a link. The script writes
pr.json, threads.json and reviews.md, and prints N, BASE_REF, HEAD_SHA and DOCS_PRS for
the rest:
node .agents/skills/tomtom-mcp-review/scripts/fetchPr.mjs "$TARGET" "$REVIEW_DIR"
DOCS_PRS lists the docs-portal PRs on the description's Docs portal: line
(CODING_GUIDELINES.md § 10); one cited anywhere else is evidence, not a
sibling. Each gets docs-<N>/ with the same three files, its diff.patch, and tree/, a sparse checkout of
the MCP pages at its head. With no access, docs-<N>/unavailable.txt says why and the review goes on
without it. In local mode, look the sibling up by its branch and fetch it alone:
gh pr list -R tomtom-internal/devportal-documentation --head tomtom-maps-mcp/<branch> --json number
node .agents/skills/tomtom-mcp-review/scripts/fetchPr.mjs --docs <number> "$REVIEW_DIR"
git fetch origin "${BASE_REF#origin/}" "pull/$N/head"
# Local mode: BASE_REF=${base:-origin/main}; HEAD_SHA=$(git rev-parse <branch>); git fetch origin main
BASE=$(git merge-base "$BASE_REF" "$HEAD_SHA")
git diff -U3 "$BASE" "$HEAD_SHA" > "$REVIEW_DIR/diff.patch"
git worktree add --detach "$REVIEW_DIR/tree" "$HEAD_SHA" # skip when reviewing in place
- The brief is copied because a checkout, stash or branch switch in your tree during the review would otherwise take it away from a reviewer still reading it.
pr.jsoncarries the title and description, which the reviewer holds the diff to.threads.jsonandreviews.mdkeep a rerun from repeating itself: each review thread with its first comment and its latest three, where a "fixed in" reply lands, and the bodies of earlier agent reviews, which hold the findings no line could anchor.pull/$N/headfetches a PR from a fork as readily as one from a branch here.- Base of a stacked PR is its
baseRefName, another feature branch, so the diff is that PR's own slice. For a local branch in a stack, the user names--base. - Reviewing in place, diff against the working tree instead (
git diff -U3 "$BASE") and listgit ls-files --others --exclude-standardinto$REVIEW_DIR/untracked.txt: the subagent reads those in full. - Whole files diff from the empty tree, so every line is a changed line, generated and vendored files
left out. Write the paths in place of
.when the user named some:
What ships with it
3 files 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.
- 7d ago Changed · +55 lines · -229 tokens per session ce5b018e6371
- 10d ago Changed · +32 lines · +33 tokens per session 24172abb4279
- 14d ago First seen · 147 lines · 196 tokens per session scan C 2c5c3e93b396
tomtom-mcp-review is a skill published in the GitHub repository tomtom-international/tomtom-maps-mcp (59 stars, last pushed yesterday), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 3,865 tokens. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-15.
Other skills, from other repositories
pr
Draft a PR body from OMC's paper trail — the smallest visual that proves the change, evidence consumed from the verify protocol (never re-collected), reversibility from the ADR test on the diff, glossary language throughout.
merge-batch
Prepare and merge an owner-selected PR batch with contributor credit and exact-candidate admission. Use for a pasted PR list; combined focused validation is diagnostic, and every actual landing still requires its applicable checks and authorization.
otto-pr-review-loop
Run Polly and OCR on a PR until all findings settle.
engram-branch-pr
PR creation workflow for Engram following the issue-first enforcement system. Trigger: When creating a pull request, opening a PR, or preparing changes for review.
github-contributor
A step-by-step guide for contributing code, tests, documentation, or build changes to open-source projects you do not maintain. It explains how to prepare a pull request that follows the project’s rules and is easier for maintainers to review.
revdiff
Review diffs, files, and documents with inline annotations in a TUI overlay, or answer questions about revdiff usage, configuration, themes, and keybindings. Opens revdiff in agterm/tmux/zellij/herdr/kitty/wezterm/cmux/ghostty/iterm2/emacs-vterm, captures annotations, and addresses them. Works in git, hg, and jj repos…