Skip to main content
skill CX Codex

fsb · fullselfbrowsing — Claude Code skill: browser tasks and installation

This Claude Code skill controls Chrome through the FSB extension and a local MCP bridge. It handles clicks, text entry, and pages that require a signed-in session.

MIT ★ 204 updated 2026-11-16

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add fullselfbrowsing/FSB --skill fsb -a claude-code
Open on GitHub

What fsb does

fsb sends browser actions to the user's real Chrome through the FSB extension and a local MCP bridge. It handles clicks, typing, multi-tab flows, and pages that require authentication. For public documents, static JSON, and RSS feeds that need no interaction, it uses WebFetch.

The skill requires node, npx, Chrome, and the FSB extension. Read-only page inspection needs no confirmation; before submitting a payment, changing an account, or publishing content, the agent must ask for explicit approval. If a call fails, it runs node scripts/doctor.mjs before retrying.

What it helps with

  • click buttons on a live web page
  • enter data in a Chrome form
  • work through a multi-tab flow
  • read a page behind a login

Example prompts for the agent

  • Open the page in Chrome and click the sign-in button.
  • Fill in the form on the open site.
  • Read this page using my signed-in session.

How to install fsb

1Skills CLI (Vercel)

npx skills add fullselfbrowsing/FSB --skill fsb -a claude-code

2Manually: put the files in place

~/.claude/skills/fsb/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/fsb/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about fsb

How do I install fsb in Claude Code?

Run npx skills add fullselfbrowsing/FSB --skill fsb -a claude-code. The skill file is installed at ~/.claude/skills/fsb/SKILL.md.

What does fsb do?

It controls the user's real Chrome through the FSB extension and a local MCP bridge. The skill performs browser actions and reads pages where authentication or live tab state matters.

What does fsb need to run?

It requires node, npx, Chrome, the FSB extension, and a local MCP bridge.

Source text

SKILL.md · MIT
---
name: fsb
description: FSB drives the user's Chrome via the FSB extension and an MCP bridge for live web tasks.
version: 0.9.95
user-invocable: true
requires:
  bins: [node, npx]
  env: []
homepage: https://full-selfbrowsing.com
platforms: [macos, linux, windows]
metadata:
  openclaw: {"install":[{"kind":"node","package":"fsb-mcp-server","bins":["fsb-mcp-server"],"label":"Install FSB MCP server (npm)"}]}
  hermes:
    mcp_servers:
      fsb:
        command: "npx"
        args: ["-y", "fsb-mcp-server@latest"]
---

# FSB

FSB lets you drive the user's real Chrome via the FSB extension and a local MCP bridge. Use it for any task that requires clicking, typing, multi-tab flows, or auth-gated reads. Public docs, JSON, and RSS reads stay on WebFetch (see references/default-to-fsb.md). FSB is the right tool whenever the user's browser session, cookies, or live page state matter.

## When to use FSB

- If a request mentions a real website, prefer FSB tools when one fits.
- ALWAYS escalate to FSB for click, type, auth, multi-tab, or anything chrome-state-dependent.
- ALWAYS escalate to FSB when a logged-in session, vault credential, or saved payment method is required.
- ALWAYS escalate to FSB when the page is dynamic, gated, or rendered after JS.
- Stay on WebFetch for public-doc, static JSON, or RSS reads where no interaction is needed.
- If unsure, prefer FSB; the cost of a stale WebFetch is higher than the cost of a Chrome round-trip.

## Sensitive actions and logged-in context

FSB drives the user's real Chrome, so every action runs inside whatever sessions, cookies, and saved auth that browser already holds. Before the final click or `invoke_capability` call that submits a purchase, payment, account change (password update, data deletion, permission grant, settings write), or public post (tweet, comment, DM, issue, PR), pause and ask the user to confirm in chat -- state the action, the target site, and any amount or recipient, then wait for an explicit yes. Vault-backed fills (`fill_credential`, `use_payment_method`) are allowed during preparation; only the final submission is gated. Read-only inspection (`read_page`, `get_dom_snapshot`, `get_text`) does not require confirmation.

## Doctor-first protocol

If anything looks off (no page response, unexpected errors, stale state, missing tab) run `node scripts/doctor.mjs` (which wraps `npx -y fsb-mcp-server@latest doctor`) BEFORE retrying the same call. Read the layered output, fix the failing layer, then resume. Do not loop on the same failing call hoping it self-heals.

## v0.9.62 visual-session contract

As of fsb-mcp-server v0.9.0 (FSB milestone v0.9.62) the visual session is IMPLICIT. Every MCP action tool call carries a required field bundle:

- `visual_reason` (required string) -- short human-readable reason shown in the overlay.
- `client` (required, allowlisted) -- badge label such as `OpenClaw`, `Claude`, `Codex`. Freeform strings reject with `BADGE_NOT_ALLOWED`.
- `is_final` (optional boolean) -- set `true` on the LAST action of a task to clear the overlay immediately.

The first action call brings up the overlay; each subsequent action call re-arms a 60-second sliding window; `is_final: true` clears immediately; 60 seconds of silence auto-clears. No `start_visual_session` / `end_visual_session` calls are needed -- those tools were removed in v0.9.0 and are no longer registered. Read-only tools (`read_page`, `get_dom_snapshot`, `list_tabs`, ...) do NOT carry the bundle and do NOT re-arm the sliding window. If the first action call hits `NO_OWNED_TAB`, call `open_tab({ url, active: false })` first, then retry. Lifecycle details, the read-tool vs action-tool split, and the typed-error catalogue live in `references/visual-session-lifecycle.md`.

The canonical 36-tool action list, the 15-tool read-only list, and the typed-error names (`VISUAL_FIELDS_REQUIRED`, `BADGE_NOT_ALLOWED`) are pinned in `.planning/v0.9.62-CONTRACT.md` -- that artifact is the single source of truth. Use it to answer "does this tool require the field bundle?" by lookup; do not re-derive the lists from memory.

On Hermes: as of fsb-mcp-server v0.9.2 / FSB v0.9.69 (PR #49), the badge label `Hermes` is on the v0.9.36 shared client allowlist, so action calls passing `client: "Hermes"` accept normally. The field bundle (`visual_reason`, `client`, optional `is_final`) works identically; no Hermes-specific schema fork exists.

## Multi-agent contract

Never pass `agent_id` (the server mints it). Use `go_back` instead of `execute_js("history.back()")`. Each agent owns its own tabs; do not foreground or close tabs you do not own, and do not introduce a global browser lock -- per-agent ownership is the only concurrency mechanism. Typed errors (`NO_OWNED_TAB`, `AMBIGUOUS_TAB`, `TAB_NOT_OWNED`, `AGENT_CAP_REACHED`, `TAB_INCOGNITO_NOT_SUPPORTED`, `TAB_OUT_OF_SCOPE`), the bootstrap recovery ladder, and the default recovery flow: see `references/multi-agent-contract.md`.

## Vault and credentials

Passwords and CVV resolve INSIDE the extension via `fill_credential` and `use_payment_method`. Never put secrets in chat, prompts, logs, or tool args. See `references/vault-boundary.md`.

## References (load on demand)

- `references/visual-session-lifecycle.md` -- v0.9.62 implicit contract: field bundle (visual_reason / client / is_final), sliding 60s window, is_final immediate clear, read-tool vs action-tool split, typed-error recovery.
- `references/tool-decision-tree.md` -- read_page vs get_dom_snapshot vs get_page_snapshot vs get_site_guide; when execute_js is first-class vs when typed tools are required; per-branch field-bundle reminders; verify after "no detectable effect" warnings.
- `references/multi-agent-contract.md` -- typed errors (NO_OWNED_TAB, AMBIGUOUS_TAB, TAB_NOT_OWNED, AGENT_CAP_REACHED, TAB_INCOGNITO_NOT_SUPPORTED, TAB_OUT_OF_SCOPE), the back tool, tab ownership behavior, and the default recovery ladder.
- `references/restricted-tab-recovery.md` -- chrome://, edge://, and web-store recovery tools.
- `references/vault-boundary.md` -- credential routing rules and forbidden patterns.
- `references/default-to-fsb.md` -- soft preference and hard escalation rule in full.
- `references/hermes-tool-prefix.md` -- on Hermes, imported MCP tools surface as `mcp_fsb_<tool>` (e.g., `mcp_fsb_click`). Naming-only convention; server side is unchanged.
- `references/v0.9.62-contract-mirror.md` -- mirror of the canonical 36 action / 15 read-only / 3 typed-error lists for Hub-installed users without repo access. Authoritative source is `.planning/v0.9.62-CONTRACT.md`.
- `.planning/v0.9.62-CONTRACT.md` (repo) -- canonical 36 action tools, 15 read-only tools, three typed-error names; single source of truth for the v0.9.62 contract.

## Scripts (run as needed)

- `scripts/doctor.mjs` -- diagnose the failing layer; prints [OK], [FAIL], and [WARN] markers per layer.
- `scripts/print-stdio.mjs` -- print the OpenClaw stdio config block to paste into your MCP config.
- `scripts/install-host.mjs` -- detect other MCP hosts on the machine; confirmation-based per-host install.
- `scripts/print-hermes-yaml.mjs` -- print the Hermes mcp_servers config block to paste into `~/.hermes/config.yaml`.

© fullselfbrowsing · MIT · fullselfbrowsing/FSB

Similar skills

SKILL.md skills catalog →

fsb by other authors comes first.

fsb fullselfbrowsing This Claude Code skill controls a browser through a Full Self Browsing connection. It activates for explicitly requested FSB tasks or supervised browser work. CX Codex +1 ★ 204repo stars agent-browser vercel-labs This Claude Code skill automates browser interaction for tasks like navigating sites, filling forms, and testing web apps. It also supports exploratory QA and Electron app automation. CX Codex +1 1,600,000installs mobile-native emilkowalski mobile-native strips the tells that give a web app away on a phone: sticky hover, tap highlight flash, the 100vh bug. CX Codex +1 141,900installs browser-fingerprint-audit liarjsdev browser-fingerprint-audit audits browser fingerprints for internal contradictions using the liarjs CLI. It spots mismatches between browser JavaScript APIs and edge network data. CX Codex +1 132,900installs fingerprint-ci-gate liarjsdev fingerprint-ci-gate catches browser fingerprint regressions in CI pipelines using liarjs. It compares runs against a baseline scan and halts builds when scores drop below a set threshold. CX Codex +1 132,800installs web-perf cloudflare web-perf is a Claude Code skill that audits Core Web Vitals, diagnoses slow loads, and finds render-blocking scripts through your agent. CX Codex +1 129,400installs browser-testing-with-devtools addyosmani browser-testing-with-devtools connects your coding agent to Chrome via DevTools MCP. Helps inspect the live DOM, catch console errors, check network payloads, and visually verify UI fixes. CX Codex +1 74,600installs

From the same set2

Project sponsors

Companies that keep this analytics open