bb is an agentic IDE that lets users control, customize, and automate a software-development environment through a desktop app, web app, command-line interface, or HTTP API. Developers use it to run work in observable threads that can be steered or handed off to other agents. Its catalogue add-ons provide skills and instructions for working with bb.
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 get-bb/bb --skill create-verification-skillgit clone --depth 1 https://github.com/get-bb/bbWrote 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/get-bb/bb/create-verification-skill)<a href="https://agentmods.dev/skills/get-bb/bb/create-verification-skill"><img src="https://agentmods.dev/badge/skills/get-bb/bb/create-verification-skill/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/get-bb/bb/create-verification-skill"><img src="https://agentmods.dev/badge/skills/get-bb/bb/create-verification-skill.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00057 | $0.01088 |
| Opus 5 | $0.00028 | $0.00544 |
| Sonnet 5 | $0.00011 | $0.00218 |
| Haiku 4.5 | $0.00006 | $0.00109 |
Grade A, and why
create-verification-skill 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 — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create a verification skill
Generate .bb/skills/verify-<app>/SKILL.md for an agent arriving cold in the
repository. Adapt the recipe to the actual app and prove it by running it.
Discover
Read repository instructions, startup scripts, routes, existing tests, and QA
docs. Establish the user surfaces, launch command, readiness signal, required
runtime, authentication, fixtures, automation harness, and observable side
effects. Prefer the repository's harness. For browser interaction without an
existing project harness, use dev-browser@next. Install it with
npm install -g dev-browser@next, then read dev-browser --help. It provides
Puppeteer scripts, named pages, page.snapshot(), and snapshot-ref selectors.
If it reports missing Chrome, run dev-browser install. Record the installed
version with the evidence; the next tag can change.
Inspect an existing verification skill before creating another. Extend it when it already owns the same app. Ask only for decisions the repository cannot answer. Do not execute this workflow for a request that only asks what a skill does or asks to verify one existing feature.
Establish isolation before launch: ports, database, host connections, browser profile, and process ownership. Use synthetic data. Do not copy a live store or reuse the user's logged-in app. Respect project rules for imported data and network-facing plugins. A new data-directory name alone does not prove that a launcher will not import existing data.
Check the real startup path. Resolve routine local setup failures within the task's scope. Report product defects and unavailable prerequisites precisely; do not change product behavior merely to make verification pass.
Generate
Write the skill with valid BB frontmatter and these sections:
- Launch: exact commands, runtime, readiness check, target discovery, and isolation. Derive ports and IDs from real output instead of guessed values.
- Doctor: a read-only check of instance identity, process ownership, build/source revision, API health, and any feature-specific prerequisite.
- Drive: documented harness commands and stable selectors from the real app. Include how CLI/API calls target the same instance as the browser.
- Evidence: capture the starting state, user action, resulting state, and relevant persisted side effect. Record the source commit, environment, command, and result. Keep evidence outside disposable runtime state.
- Cleanup: stop only processes and browser sessions this run owns. Verify they stopped and evidence survives. Handle failed attempts too.
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 · 99 lines · 57 tokens per session scan A 2d6685b47356
create-verification-skill is a skill published in the GitHub repository get-bb/bb (3,416 stars, last pushed today), licensed MIT. It adds 57 tokens to every session and 1,088 once invoked, about $0.0003 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 2026-09-07.
Other skills, from other repositories
browser-qa
Use when you need lightweight browser QA for a web page, local HTML file, or app: inspect console errors, broken assets, keyboard/focus behavior, viewport readability, and publish evidence-backed findings JSON through a local HTML report viewer.
playwright-explore-website
A browser-testing workflow that explores a website through Playwright, a tool for controlling a web browser with code. It records important user flows, page elements, expected results, and possible test cases.
vite-browser-release-smoke
Pre-release smoke verification workflow for Vite apps with vite-browser. Use this whenever users ask for final checks before merge/release, including core path validation, runtime sanity, HMR diagnosis, and evidence-based sign-off.
playwright-best-practices
A set of guidelines for writing and maintaining Playwright tests for websites and applications. It covers browser-based end-to-end tests, API tests, component tests, and specialised cases such as authentication or file uploads.
webapp-testing
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
playwright-cli
Automate browser interactions, test web pages and work with Playwright tests.