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/get-bb/bb/inline-visnpx skills add get-bb/bb --skill inline-visgit clone --depth 1 https://github.com/get-bb/bbWhat 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 | $0.00045 | $0.00485 |
| Opus 5 | $0.00023 | $0.00243 |
| Sonnet 5 | $0.00009 | $0.00097 |
| Haiku 4.5 | $0.00005 | $0.00049 |
Grade A, and why
inline-vis 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 2d 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.
What it actually says
Inline HTML visualizations
When the user should see a small HTML demo, chart, or report inline in the
assistant message, write (or update) a workspace-relative .html file, then
emit this message directive as its own block (not inside a fenced code
block):
::inline-vis{file="demo.html"}
Rules
fileis workspace-relative (e.g.demo.html,charts/out.html). Never use absolute paths.heightis optional and sets the iframe viewport height in pixels. It must be a whole number from 120 through 1200; omit it for the 224px default.- Only
.html/.htmfiles are accepted. - Inline and external CSS/JavaScript are supported. Remote images, fonts, media, fetches, and WebSockets are also allowed subject to normal browser CORS, mixed-content, and remote-server policies. Scripts execute in an opaque-origin iframe and cannot access the bb page, cookies, or storage.
- Keep files small (under the sidebar preview's 5 MiB document limit).
- Emit the directive only after the file exists on disk in the current thread workspace.
- Do not put the directive inside backticks or a markdown code fence, or it stays literal text.
- Incomplete streaming syntax stays literal until the closing
}arrives — emit a complete directive in one piece when possible.
Example turn
- Write
demo.htmlwith a simple visualization (inline styles, no remote JS). - Reply with a short caption and the directive:
Here is the chart:
::inline-vis{file="demo.html"}
For a taller visualization, emit for example:
::inline-vis{file="demo.html" height="640"}
The bb app replaces the directive with a sandboxed preview. If the plugin is disabled or the path is invalid, users see the original directive source or an inline error from the plugin.
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.
- 2d ago First seen · 56 lines · 45 tokens per session scan A 0722c2d286c2
inline-vis is a skill published in the GitHub repository get-bb/bb (2,762 stars, last pushed 2d ago), licensed MIT. It adds 45 tokens to every session and 485 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
cradle-plugin-browser-use
Control Cradle's in-app browser via MCP tools. Use when the Claude Agent needs to navigate websites, click elements, type text, take screenshots, read page content, or inspect DOM structure within the Cradle desktop app's embedded browser panel. Triggers on "open URL", "navigate to", "click the button", "fill the…
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…
using-ao
Catalog of the AO (Agent Orchestrator) ao CLI: spawning workers, managing sessions and projects, sending messages, controlling the shared browser, previewing pages, and daemon control. Use when using the ao CLI, spawning workers, or managing AO sessions in an AO workspace.
d3k
Use when the user asks to use d3k, run/dev/test/debug a web project with d3k, or reproduce a browser issue. Own the runtime: reuse or background-start d3k non-interactively, wait for readiness, use its project-stable managed Chrome profile, and inspect unified browser/server evidence.
ui-dev
This skill MUST be used whenever the task involves UI development, renderer code changes, adding or modifying components, creating modals or dialogs, working with CSS styles, building new UI features, or touching any file in src/renderer/. Use this skill when the user asks to "add a button", "create a modal", "add a…
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.