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/patrickdappollonio/claude-plugins/visual-recapnpx skills add patrickdappollonio/claude-plugins --skill visual-recapgit clone --depth 1 https://github.com/patrickdappollonio/claude-pluginsWrote 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/patrickdappollonio/claude-plugins/visual-recap)<a href="https://agentmods.dev/skills/patrickdappollonio/claude-plugins/visual-recap"><img src="https://agentmods.dev/badge/skills/patrickdappollonio/claude-plugins/visual-recap.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 | $0.00112 | $0.04433 |
| Opus 5 | $0.00056 | $0.02217 |
| Sonnet 5 | $0.00022 | $0.00887 |
| Haiku 4.5 | $0.00011 | $0.00443 |
Grade A, and why
visual-recap 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 3d 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 — 321 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Visual Recap
Turn a change — a PR, branch, commit range, or the working tree — into an interactive review document served entirely from the user's machine.
Before anything else, fix who you're writing for: the CEO of the company — a tech-savvy non-developer. Not a fellow engineer, not the person who wrote the code, and not someone who will ever open the repo. This shapes every sentence you write. They want the business logic — what changed, why it matters, what to watch — so explain behavior in plain language and reach for code only when the reader must see it to understand. The linter warns when plain-language sections name code symbols, and those findings must be fixed like any other.
Acknowledge first, then work quietly. Before you do anything else, reply with one short sentence that acknowledges the request and says you're gathering what you need — e.g. "On it — let me pull the diff together and build a visual recap of PR 142." This is the one message the user should get up front; never jump straight into tool calls with no reply. Then spend your tokens on the document, not on narrating: move through steps 1–4 without a play-by-play — no "Step 1: capturing the diff…", no restating the captured diff or the inventory, no "here's what I found." Your next message after the acknowledgement is the link (once served) with a one-line pointer. Every token you'd spend describing the work, spend instead making the document more complete.
The sequence — do every step, in order; the last two are the ones agents skip:
- Capture the diff (once).
- Inventory it (silently).
- Write the recap file.
- Lint it and fix every finding — required, not optional (§3).
- Self-review the rendered document — required (§3): re-read it top to bottom against your inventory before anyone else sees it.
- Serve it and hand over the link.
- Read and act on comments — then back to 4 and 5 for every edit.
Every write runs write → lint → verify — the first draft and every revision. After an edit, do not serve, share, or say "done" until the linter is clean on that file and you have re-read the changed sections against its closing reminder. Revisions are where agents skip this.
What ships with it
31 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.
- server/assets/app.css 57 KB
- server/assets/app.js 162 KB runs code
- server/assets/index.html 2.6 KB
- server/assets/vendor/diff2html.min.css 17 KB
- server/assets/vendor/diff2html.min.js 66 KB runs code
- server/assets/vendor/graphre.js 38 KB runs code
- server/assets/vendor/highlight.min.js 119 KB runs code
- server/assets/vendor/hljs-github-dark.min.css 1.3 KB
- server/assets/vendor/hljs-github.min.css 1.3 KB
- server/assets/vendor/htm.umd.js 1.3 KB runs code
- server/assets/vendor/js-yaml.min.js 39 KB runs code
- server/assets/vendor/manifest.json 5.6 KB
- server/assets/vendor/marked.min.js 35 KB runs code
- server/assets/vendor/mermaid.min.js 3258 KB runs code
- server/assets/vendor/nomnoml.js 71 KB runs code
- server/assets/vendor/preact-hooks.umd.js 3.8 KB runs code
- server/assets/vendor/preact.min.js 11 KB runs code
- server/assets/vendor/purify.min.js 21 KB runs code
- server/bin/visual-docs-lint.js 22 KB runs code
- server/bin/visual-docs-server.js 26 KB runs code
- server/examples/demo-plan.md 7.3 KB
- server/lib/export.js 7.1 KB runs code
- server/lib/prefs.js 3.3 KB runs code
- server/lib/server.js 51 KB runs code
- server/lib/trash.js 5.1 KB runs code
- server/lib/version.js 1.8 KB runs code
- server/package.json 1.1 KB
- server/README.md 7.6 KB
- server/scripts/update-vendor.mjs 11 KB runs code
- shared/authoring-guide.md 22 KB
- shared/document-quality.md 15 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.
- 3d ago First seen · 321 lines · 112 tokens per session scan A ea6a0ac26c1e
visual-recap is a skill published in the GitHub repository patrickdappollonio/claude-plugins (9 stars, last pushed 4d ago), licensed MIT. It adds 112 tokens to every session and 4,433 once invoked, about $0.0006 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…