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/m4ttstack/fast-browser/reviewing-flowsnpx skills add m4ttstack/fast-browser --skill reviewing-flowsgit clone --depth 1 https://github.com/m4ttstack/fast-browserWhat 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 | $0.00046 | $0.01642 |
| Opus 5 | $0.00023 | $0.00821 |
| Sonnet 5 | $0.00009 | $0.00328 |
| Haiku 4.5 | $0.00005 | $0.00164 |
Grade A, and why
reviewing-flows 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 — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reviewing Flows
fast-browser flows approve shows a step count, not steps. A human is asked to
trust actions they cannot read. This skill reads the queue, clears what can
never work, and puts what is left in front of them with its steps rendered.
What this skill may and may not do
This skill may reject. It may never approve.
Approval needs a TTY this skill does not have, and that is the gate working as designed, not an obstacle to route around. Never wrap, pipe, script, or otherwise automate the confirm prompt, and never simulate the human's keystrokes with a terminal-automation tool. Print the command for the human to run themselves; do not run it yourself.
Everything here moves flows toward less runnable, never more. That is what makes acting without asking safe.
Phase 1: survey
Get the inventory:
fast-browser flows list --json
That gives tier, name, description, origin, health, lastHealed, and
warnings. It does not carry steps, side effects, or args, so read the
artifacts directly for those. Pending and ready live in separate directories:
~/.fast-browser/flows-pending/<name>.flow.json
~/.fast-browser/flows/<name>.flow.json
Only pending flows get bucketed, but bucket 3 below compares against ready flows too, so read both.
Sort every pending flow into exactly one bucket, first match wins:
unapprovable... any step hasop: "js". Approval can never make it runnable; it needs re-recording.dead-origin... the origin's host islocalhost,127.0.0.1, or::1, and the origin is not on the user's keep-list. Test the host alone, ignoring the port, but keep-list entries are whole origins, port included:http://localhost:6001can be kept without keepinglocalhost:7930.superseded... another flow on the same origin covers this one and wins the tiebreak. B covers A when A's ordered sequence of(op, target role, target name)triples appears as a subsequence of B's. B wins the tiebreak when it has more steps, or the same number of steps and a name that sorts first in plain byte order.
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.
- yesterday First seen · 162 lines · 46 tokens per session scan A eb243bbee81e
reviewing-flows is a skill published in the GitHub repository m4ttstack/fast-browser (0 stars, last pushed 7d ago), licensed MIT. It adds 46 tokens to every session and 1,642 once invoked, about $0.0002 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
webcmd-browser
Use when a live browser task requires Playwright interaction, authenticated handoff, visible UI verification, or ad-hoc page inspection.
moli-webfetch
Fetch, inspect, crawl, and capture live, JavaScript-rendered websites with Moli. Use when Codex needs current web content, web research, fact lookup, link following, a bounded crawl, client-rendered or response-gated content, network diagnostics, or a standalone HTML, Markdown, JSON, semantic-tree, viewport or…
moli-cdp-server
Start Moli's CDP server and connect Playwright, Puppeteer, or raw CDP clients. Use to run a headless-browser CDP endpoint, replace a Chromium process, attach over CDP, enable real layout and screenshot surfaces, or diagnose CDP discovery, connection, and target startup—even when Moli is not named.
surf
Control Chrome browser via CLI for testing, automation, and debugging. Use when the user needs browser automation, screenshots, form filling, page inspection, network/CPU emulation, DevTools streaming, or AI queries via ChatGPT/Gemini/Perplexity/Grok/AI Studio.
surf-codebase
Navigate and modify surf-cli codebase - Chrome extension + native host for AI browser automation. Use for surf-cli code work, architecture questions, implementing browser control/CDP/accessibility/network features.
playwright-rs-usage
Procedural reference for using playwright-rs in Rust browser-automation code — object model (Browser/Context/Page/Locator), the locator!() macro, builder pattern for options, auto-wait semantics, adding the crate and installing its browsers, and how to capture / inspect traces for failure diagnosis. Use when writing…