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 agentmods add skills/indyraider/annotate-plugin/annotatenpx skills add indyraider/annotate-plugin --skill annotategit clone --depth 1 https://github.com/indyraider/annotate-pluginWhat 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 | $0.00148 | $0.09907 |
| Opus 5 | $0.00074 | $0.04953 |
| Sonnet 5 | $0.00030 | $0.01981 |
| Haiku 4.5 | $0.00015 | $0.00991 |
Grade A, and why
annotate 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 yesterday.
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 — 602 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/annotate — point-and-comment on the live app
Matt marks up the running app visually; the comments flow back here to be fixed.
This skill is the loop you run. The implementation is the eight modules in overlay/,
injected into the page from disk by the Playwright process — so nothing is pasted through
your context and nothing is fetched over the network.
Setup
-
URL: use the
[url]arg, elsehttp://localhost:3000. Assume the dev server is already up (Matt keeps:3000running — never kill it). If it's not, ask. If the[url]arg is a remote address (e.g./annotate https://someothersite.com), there is no local dev server to wait for — that's a normal invocation, most often heading for Study mode (see below), which is built to work against any site, not just the local app. -
Open the page:
browser_navigateto the URL (reuse the existing browser). -
Boot the overlay — one call, no server, no pasted source:
browser_run_code_unsafe({ code: `async (page) => { const FILES = ["core.js","palette.js","ui.js","point.js","measure.js","study-motion.js","study.js","index.js"]; for (const f of FILES) await page.context().addInitScript({ path: "<this skill's directory>/overlay/" + f }); await page.reload({ waitUntil: "domcontentloaded" }); return await page.evaluate(() => { window.__annotatorMods.index.setup(); return "ready"; }); }` })addInitScripttakes a path, so Playwright reads the files in its own process and injects them over CDP before the document's own scripts — the same mechanism asbrowser_evaluate, which Chromium does not subject to page CSP. Verified booting ongithub.comunderdefault-src 'none', and onstripe.com,linear.appand localhost. Order matters: it is a dependency chain, andoverlay.test.cjsasserts this list matchescore.js'sOWN_MODULE_FILESand the directory itself.What NOT to try — all four fail, and are recorded so nobody loses an afternoon rediscovering them.
browser_run_code_unsaferuns the snippet in a barevmsandbox whose only host object ispage: norequire("require is not defined"), no dynamicimport("a dynamic import callback was not specified"), andpage.addScriptTag({ path })is refused by CSP both with and without a CDPPage.setBypassCSP— the bypass does not apply to a document that has already loaded.
What ships with it
10 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.
- overlay.test.cjs 80 KB runs code
- overlay/core.js 29 KB runs code
- overlay/index.js 15 KB runs code
- overlay/measure.js 9.9 KB runs code
- overlay/palette.js 3.5 KB runs code
- overlay/point.js 16 KB runs code
- overlay/study-motion.js 18 KB runs code
- overlay/study.js 25 KB runs code
- overlay/ui.js 23 KB runs code
- templates/design-language.md 2.5 KB
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.
- yesterday First seen · 602 lines · 148 tokens per session scan A 8fe530c30a7e
annotate is a skill published in the GitHub repository indyraider/annotate-plugin (2 stars, last pushed 11d ago), licensed MIT. It adds 148 tokens to every session and 9,907 once invoked, about $0.0007 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
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
use-agent-browser-for-airi
Test AIRI display-model imports with agent-browser across stage-tamagotchi Electron, stage-web, and stage-pocket mobile web layouts. Use when uploading and verifying contributor-supplied Live2D ZIP, VRM, or MMD ZIP/PMX/PMD files through AIRI's model selector, including onboarding bypass, format-specific import…
opencli-sitemap-author
Use when creating or maintaining OpenCLI site sitemaps: agent-facing navigation, page-state, action, workflow, API-reference, pitfall, and fallback knowledge for a website. Use after browser exploration discovers durable site context, when a sitemap is stale, or when promoting local site knowledge into the repo.
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.
mintlify-preview
Run the public Mintlify product docs site locally for live preview. Use when previewing or iterating on docs under docs/ (.mdx/.md pages, docs.json nav), or when the user says "start the docs", "run mintlify", "preview the docs site".