Skip to content
Skills Library
Documents & office · Conversion & formats

ak-render

/ak-render

Build a polished offline HTML report, dashboard or architecture review page with AK Render. Write a short YAML Page Spec, validate it, and compile it to one self-contained HTML file instead of hand-writing HTML, CSS or JavaScript. Use when the user asks for an HTML report, a status or incident report, a KPI dashboard, an architecture or design review page, or a shareable page that presents results already in the conversation. Needs Node.js 20.11+ on the machine running Claude; not for images, video, audio or slide decks.

Repository
ak-render ↗
Licence
MIT
Last repository change
08.11.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
4
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 08.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Build a polished offline HTML report, dashboard or architecture review page with AK Render. Write a short YAML Page Spec, validate it, and compile it to one self-contained HTML file instead of hand-writing HTML, CSS or JavaScript. Use when the user asks for an HTML report, a status or incident report, a KPI dashboard, an architecture or design review page, or a shareable page that presents results already in the conversation. Needs Node.js 20.11+ on the machine running Claude; not for images, video, audio or slide decks.

Original author description

Build a polished offline HTML report, dashboard or architecture review page with AK Render. Write a short YAML Page Spec, validate it, and compile it to one self-contained HTML file instead of hand-writing HTML, CSS or JavaScript. Use when the user asks for an HTML report, a status or incident report, a KPI dashboard, an architecture or design review page, or a shareable page that presents results already in the conversation. Needs Node.js 20.11+ on the machine running Claude; not for images, video, audio or slide decks.

How it works

  1. Tools
  2. The loop: brief → spec → validate → render → inspect
  3. Rules
  4. Minimal spec
Original SKILL.md headings · 118 lines, 692 words
SKILL.md preview
---
name: ak-render
description: Build a polished offline HTML report, dashboard or architecture review page with AK Render. Write a short YAML Page Spec, validate it, and compile it to one self-contained HTML file instead of hand-writing HTML, CSS or JavaScript. Use when the user asks for an HTML report, a status or incident report, a KPI dashboard, an architecture or design review page, or a shareable page that presents results already in the conversation. Needs Node.js 20.11+ on the machine running Claude; not for images, video, audio or slide decks.
---

# AK Render

AK Render compiles a YAML or JSON **Page Spec** into one deterministic,
self-contained HTML file. The page opens from `file://` and makes no network
request. You describe what the page says; the compiler owns layout, colour,
typography, accessibility and security.

This plugin pins `@bestagentkits/render@0.3.0`. Everything below works with
that version; don't use commands that it does not list in `--help`.

## Tools

When the `ak-render` MCP tools are available (`catalog`, `search-catalog`,
`describe`, `recipes`, `recipe`, `validate`, `render`, `themes`), use them.
`spec` may be YAML or JSON text. `render` writes the HTML to `out` and returns
only a summary.

Otherwise use the CLI with the same pinned version. Every command except
`mcp` accepts `--json`, and none of them prompts:

```bash
npx -y @bestagentkits/render@0.3.0 --version
```

If `npx` is missing or Node is older than 20.11, stop and tell the user to
install Node.js 20.11 or later. Don't fall back to hand-written HTML.

## The loop: brief → spec → validate → render → inspect

1. **Pick a starting point.** For a report, dashboard or architecture review,
   start from a recipe or from a bundled example and replace its content:

   | Brief | Recipe | Bundled example |
   | --- | --- | --- |
   | Report, recap, postmortem | `research-report`, `incident-report`, `release-recap` | `examples/report.yaml` |
   | KPI or service dashboard | `dashboard` | `examples/dashboard.yaml` |
   | Architecture or design review | `architecture-review`, `decision-memo` | `examples/architecture.yaml` |

   ```bash
   npx -y @bestagentkits/render@0.3.0 recipes
   npx -y @bestagentkits/render@0.3.0 recipe dashboard > page.yaml
   ```

   The examples sit in the `examples/` folder next to this file. Their numbers
   are fictional sample data, so replace every value.

2. **Find and describe only the blocks you need.**

   ```bash
   npx -y @bestagentkits/render@0.3.0 search-catalog sortable table
   npx -y @bestagentkits/render@0.3.0 describe kpi chart data-table --compact
   ```

3. **Write the spec** to a `.yaml` file. Every number, date, name and finding
   must come from the conversation, the repository or tool output. If a value
   is unknown, leave the block out or ask; never fill it with a guess.

4. **Validate and fix.** Exit code 1 means fix and retry:

   ```bash
   npx -y @bestagentkits/render@0.3.0 validate page.yaml --json
   ```

   The result is `{ ok, diagnostics[] }`. Each diagnostic has a stable `code`
   and a JSON `path` such as `$.blocks[0].titl`; for an unknown prop,
   `details.allowed` lists the valid ones. Fix each diagnostic at its path and
   validate again until `ok` is `true`.

5. **Render** to an `.html` file:

   ```bash
   npx -y @bestagentkits/render@0.3.0 page.yaml --out page.html --json
   ```

   Exit code 1 means the spec is invalid and nothing was written; exit code 2
   means a file could not be read or written. Report the message to the user.

6. **Inspect the result.** Read the JSON summary, not the HTML: confirm `out`
   is the path you asked for, `title` is right and `warnings` is empty (fix the
   spec for each warning and render again). Don't read the HTML back into
   context. Give the user the file path, and open it once with `open` (macOS),
   `xdg-open` (Linux) or `start` (Windows) only if the file did not exist
   before this render; otherwise ask them to refresh.

## Rules

- **No raw HTML, CSS or JS.** The spec has no field for them; keys such as
  `rawHTML`, `iframe`, `srcdoc` or `script` are rejected. Don't post-process
  the emitted file. Links accept `https`, `http`, `mailto` and relative paths.
- **Keep the page offline.** `policy.network` defaults to `deny`; only opt in
  when the user asks for remote media.
- **Choose a theme by name.** List them with `themes` and set
  `theme.preset`; don't restyle the page.
- **Not for generated media.** AK Render lays out text, tables, charts and
  diagrams from data you already have. It does not create images, video or
  audio.

## Minimal spec

```yaml
version: 1
meta:
  title: Migration plan
  description: How we move the API to the new gateway.
blocks:
  - type: hero
    eyebrow: Plan
    title: Migration plan
    description: Move the public API to the new gateway without downtime.
  - type: steps
    items:
      - title: Mirror traffic
        text: Send a copy of production requests to the new gateway.
      - title: Switch reads
        text: Route read endpoints once error rates match.
```
Files in this package (4)
  • SKILL.md 5 KB
  • examples/architecture.yaml 3 KB
  • examples/dashboard.yaml 2 KB
  • examples/report.yaml 2 KB

Versions

RepositoryScoreLicence
This page · bestagentkits/ak-render100MIT
bestagentkits/ak-render100MIT

Database snapshot: 13.11.2026 · package version 18b3516ef4f9 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100v1-html-it
    Use when creating a self-contained HTML page, report, or interactive explainer. Triggers on "one-page dashboard", "shareable page", or "interactive report".
  • 100single-html-document
    Build polished, self-contained single-file HTML documents: one browser-openable .html with all content, CSS, JavaScript, media, and export contracts inlined. Use whenever the user wants a rich, visual, or interactive document delivered as a standalone HTML file. Triggers: 'make…
  • 95report-with-html
    Produces the reader-facing HTML artifact — report, dashboard, architecture view, data browser, or review workbench — evidence-backed, browser-verified. Use when the deliverable IS a visual HTML report, or the user asks for an HTML/web presentation. Load…
  • 100render-report
    Render existing .ux/audits reports into self-contained, offline HTML pages — findings as cards, the captured screenshots as an interactive walk-through flipbook, plus the roll-up dashboard and an index landing page. Use for "render the report to HTML", "HTML view of the audit",…
  • 100sharing-results
    The user wants to share or hand off what was found in a video — "send this analysis to my team", "make a page I can share", "export the findings", "give me something I can attach to the ticket". Use this to render a self-contained offline HTML viewer page with the frames,…
  • 100report-card
    Render a finished BS report as a self-contained HTML page — score hero, filterable claims, readable on a phone, prints to a clean PDF. Use when the user wants to open, share, print, or show a report to someone who is not going to read a markdown table, or asks for "the HTML…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.