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 skills add smallnest/goclaw --skill actionbookgit clone --depth 1 https://github.com/smallnest/goclawWrote 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/smallnest/goclaw/actionbook)<a href="https://agentmods.dev/skills/smallnest/goclaw/actionbook"><img src="https://agentmods.dev/badge/skills/smallnest/goclaw/actionbook.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 5.1 | $0.00049 | $0.03574 |
| Opus 5 | $0.00024 | $0.01787 |
| Sonnet 5 | $0.00010 | $0.00715 |
| Haiku 4.5 | $0.00005 | $0.00357 |
Grade A, and why
actionbook 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 8d 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 — 421 lines — stays where its author put it; the contents beside it link to each section on GitHub.
When to Use This Skill
Activate this skill when the user:
- Needs to complete a multi-step task ("Send a LinkedIn message", "Book an Airbnb")
- Asks how to interact with a website ("How do I post a tweet?")
- Builds browser-based AI agents or web scrapers
- Writes E2E tests for external websites
- Navigates to any new page during browser automation
- Wants to control their existing Chrome browser (Extension mode)
Browser Modes
Actionbook supports two browser control modes:
| Mode | Flag | Use Case |
|---|---|---|
| CDP (default) | (none) | Launches a dedicated browser instance via Chrome DevTools Protocol |
| Extension | --extension |
Controls the user's existing Chrome browser via a Chrome Extension + WebSocket bridge |
When to use Extension mode:
- The user wants to operate on their already-open Chrome (with existing logins, cookies, tabs)
- The task requires interacting with pages that need the user's real session state
- The user explicitly mentions their Chrome browser, extension, or existing tabs
When to use CDP mode (default):
- Clean browser environment is preferred
- Headless automation or CI/CD
- Profile-based session isolation is needed
All actionbook browser commands work identically in both modes. The only difference is adding --extension flag (or setting ACTIONBOOK_EXTENSION=1).
How to Use
Phase 1: Get Action Manual
# Step 1: Search for action manuals
actionbook search "arxiv search papers"
# Returns: area IDs with descriptions
# Step 2: Get the full manual (use area_id from search results)
actionbook get "arxiv.org:/search/advanced:default"
# Returns: Page structure, UI Elements with CSS/XPath selectors
Phase 2: Execute with Browser (CDP mode — default)
# Step 3: Open browser
actionbook browser open "https://arxiv.org/search/advanced"
# Step 4: Use CSS selectors from Action Manual directly
actionbook browser fill "#terms-0-term" "Neural Network"
actionbook browser select "#terms-0-field" "title"
actionbook browser click "#date-filter_by-2"
actionbook browser fill "#date-year" "2025"
actionbook browser click "form[action='/search/advanced'] button.is-link"
# Step 5: Wait for results
actionbook browser wait-nav
# Step 6: Extract data
actionbook browser text
# Step 7: Close browser
actionbook browser close
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.
- 8d ago First seen · 421 lines · 49 tokens per session scan A e6366f07a310
actionbook is a skill published in the GitHub repository smallnest/goclaw (603 stars, last pushed 5mo ago), licensed MIT. It adds 49 tokens to every session and 3,574 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 2026-08-30.
Other skills, from other repositories
browser-automation
Playwright-based browser automation patterns for autonomous web interaction.
pm-preview
A tool that previews Markdown product documents in a web browser and refreshes the view when the documents change. Markdown is a plain-text format commonly used for project documentation.
xhs-auto-publisher
A cloud-server workflow for preparing and publishing image-and-text posts on Xiaohongshu, a Chinese social-media platform. It uses browser automation while allowing a person to scan a login QR code.
agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a…
agent-browser
A command-line tool for controlling a web browser without a visible window. It can open pages, inspect their accessibility structure, interact with forms, take screenshots, create PDFs, and run JavaScript.
browser-control
A browser-automation guide using SeleniumBase's undetected Chrome mode. It is intended for websites that need interaction, JavaScript rendering, login steps, screenshots, or protection such as Cloudflare.