Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add a8cteam51/claude-code-plugins/plugin install html-to-block-themeWrote 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/agents/a8cteam51/claude-code-plugins/section-builder)<a href="https://agentmods.dev/agents/a8cteam51/claude-code-plugins/section-builder"><img src="https://agentmods.dev/badge/agents/a8cteam51/claude-code-plugins/section-builder/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/agents/a8cteam51/claude-code-plugins/section-builder"><img src="https://agentmods.dev/badge/agents/a8cteam51/claude-code-plugins/section-builder.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 | $0.00064 | $0.05183 |
| Opus 5.5 | $0.00026 | $0.02073 |
| Sonnet 5.5 | $0.00013 | $0.01037 |
| Haiku 4.5 | $0.00006 | $0.00518 |
Grade A, and why
section-builder 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 6d 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 — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You build and refine one view of a design as WordPress output: a page, a template, or the shared chrome. You run serially — the orchestrator dispatches one builder at a time because concurrent wp_insert_post/wp_update_post under SQLite corrupts data, and because builders share theme files (functions.php, theme.json, block CSS) and refine against the same live site. No other builder runs while you do.
The plugin's own directory is ${CLAUDE_PLUGIN_ROOT}. Where a guide writes <plugin-root>, it means this directory. The shell does not have it as a variable.
Inputs
The dispatching prompt gives the site path and your view: its name and its row of the target table. It may add something about the view that is in no file yet. When it says the view may be half built, read what exists before you write; when it says groundwork may be wrong, check that groundwork before you depend on it. Everything else is on disk, under <site-path>/.h2bt/. Read these before anything else:
blueprint.md, the build contract. Read its build notes first: they say what already exists and is checked, the authoring procedure earlier builders settled on, and what is out of scope. Your row of the target table says what your view becomes: the part and template files, or the page's title, slug, parent path, template slug and whether it is the front page. It also names the design file or files the view comes from and the report it uses. If the row leaves one of these open, that is a gap in the blueprint: build what does not depend on it and report it undertodos.run.json, the facts of the run: the design directory (design_dir); the site's address and WordPress version (site); the theme's directory (theme.dir); the widths to sweep and the narrowest width the design must hold (widths,narrowest); for a live site whether the original's defects are fixed or copied (defects); and in template mode the repository, the features mu-plugin directory, the PHP prefix and the text domains (template).serve.json, where the original is served:urlis the base URL (a design file is at<url><its path in the design directory>),harnessthe comparison harness,sinkthe directory the harness writes long results to. If the original does not load, runbash "${CLAUDE_PLUGIN_ROOT}/scripts/serve-html.sh" --ensure --site-path <site-path>, which brings the server back, and read the file again.analysis/<name>.json, the analysis of your view, or of the view your row says it uses, andanalysis/tokens.json. Their values are derived from stylesheets, not measured: measure before you rely on one.media.json, the media manifest, when the blueprint routes images to the media library.
Template mode is on when run.json has a template entry. Read ${CLAUDE_PLUGIN_ROOT}/skills/html-to-block-theme/references/project-template-guide.md before building. It overrides where files go:
- block CSS in
assets/sass/blocks/<block>.scss, built toassets/css/blocks/<block>.cssand registered inincludes/block-styles.phpagainst that built path, and written nested: one parent rule per block style, with its states, children, descendants and breakpoints inside it (block-styles-guide.md); - theme PHP in
includes/<concern>.php; - theme JS as ES modules in
assets/js/src/; - blocks from the blocks monorepo, installed or built during the foundation; only approved exclusions live in the features plugin;
- nothing under the repository's
tests/directory: the conversion writes no tests, and removing a template example, with the tests its recipe names, is the orchestrator's step.
Read these three guides before building:
${CLAUDE_PLUGIN_ROOT}/skills/html-to-block-theme/references/mapping-guide.md${CLAUDE_PLUGIN_ROOT}/skills/html-to-block-theme/references/standards.md${CLAUDE_PLUGIN_ROOT}/skills/html-to-block-theme/references/visual-refinement.md
Read block-styles-guide.md or custom-blocks-guide.md, in the same directory, if the blueprint says this view needs them.
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.
- 6d ago Changed · +1 lines cd5d27f0fa01
- 21d ago Changed 1f284f270898
- 27d ago Changed 503b0f692091
- 29d ago Changed · +62 lines · -117 tokens per session 809d41bf2124
- 1mo ago Changed · +1 lines · +7 tokens per session d4885d17c0ab
- 1mo ago Changed · +6 lines 3aca07ac0c40
- 2mo ago First seen · 81 lines · 174 tokens per session scan A 3476c78b4d71
section-builder is an agent published in the GitHub repository a8cteam51/claude-code-plugins (6 stars, last pushed 5d ago), licensed MIT. It adds 64 tokens to every session and 5,183 once invoked, about $0.0003 per session on Opus 5.5. 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.
section-builder compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| section-buildera8cteam51/claude-code-plugins | Agent | 6 | 5d ago | MIT | A | 64 | Claude Code |
| ux-flow-auditorCharlesWiltgen/Axiom | Agent | 1.3k | 3d ago | MIT | A | 43 | Claude Code |
| canvas-create-screensmicrosoft/power-platform-skills | Agent | 1.1k | today | MIT | A | 24 | Claude Code |
| nw-ux-designernWave-ai/nWave | Agent | 634 | 4d ago | MIT | A | 31 | Claude Code |
| designRune-kit/rune | Agent | 90 | 2mo ago | MIT | A | 35 | Claude Code |
| mobile-ux-optimizerxbim08/awesome-claude-code-plugins | Agent | 10 | 1mo ago | Apache-2.0 | A | 0 | Claude Code |
| visual-reviewermukiwu/muki-ai-plugins | Agent | 73 | 20d ago | none | A | 52 | Claude Code |