Skip to content
Agentmart
Submit a tool

mawf-archify

Render, validate, and deliver architecture diagrams through the MAWF-locked Archify engine.

Skill by imBlanker

IndexedMIT★ 1
Write a review

Chalk Labs

Need help deploying mawf-archify?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/imblanker-mawf-archify

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Render, validate, and deliver architecture diagrams through the MAWF-locked Archify engine. Use for authored-IR diagram work with evidence-bound validation; only small IR summaries enter agent context.

Use cases: agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/mawf-archify

What's inside

Skill name
mawf-archify
Description
Render, validate, and deliver architecture diagrams through the MAWF-locked Archify engine. Use for authored-IR diagram work with evidence-bound validation; only small IR summaries enter agent context.
License
MIT
Instructions
57 lines in SKILL.md · view on GitHub

Files (2)

  • SKILL.md
  • authoring-cheatsheet.md · references/

SKILL.md (first lines)

# MAWF Archify

Thin skill over the Archify engine locked as an MAWF component (engine commit + digest pinned in the component manifest; MIT — retains the Cocoon AI copyright notice, see `NOTICE-MAWF-INTEGRATIONS.md` at the skills root).

## Availability gate

The engine is a managed component, NOT the npm `archify` package (that name is an unrelated 2022 package). Check before use:

```bash
mawf components status archify   # once the components CLI ships
mawf archify doctor
```

If the component is missing, stop at the gate and tell the user: **requires `mawf components install archify --from-file <archive>`**. Do not improvise with an untracked engine copy, and do not substitute the DSH-specific bundle (different version, higher Node floor, different snapshot). During development, a locked engine checkout can be used via `MAWF_ARCHIFY_SRC=<checkout> mawf archify <cmd> ...` — record engine version and digest with the artifact.

## Invocation — exact locked-engine CLI (raw pass-through)

`mawf archify` forwards argv **verbatim** to the locked engine — it is not a redesigned CLI. Flags, values and order are preserved; use the engine's real grammar (excerpt from the locked commit's own usage; consult `--help` of the resolved engine for full detail):

```bash
mawf archify render    <type> <input.json> [output.html] [--quality standard|showcase] [--repo-root <path>]
mawf archify validate  <type> <input.json> [--json] [--layout-json] [--quality standard|showcase] [--repo-root <path>]
mawf archify deliver   <type> <input.json> [output.html] [--json] [--open] [--quality ...] [--repo-root <path>]
mawf archify preview   <type> <input.json> [output.html] [--no-open] [--quality ...]
mawf archify compare architecture <base.json> <head.json> [out.html] [--receipt <path>] [--json]
mawf archify check <output.html>          # artifact integrity check
mawf archify visual-check <output.html> [--json]   # requires Chrome; exit 2 = skipped
mawf archify migrate workflow <old.json> <new.json> --to-schema 2 [--json]
mawf archify inspect <type> <input.json>  # architecture/workflow layout dump
mawf archify doctor
```

`<type>` ∈ `architecture | workflow | sequence | dataflow | lifecycle` — the engine **requires** the diagram type as the first argument of render/validate/deliver/preview. JSON receipts (`--json`) are the machine-readable contract; agent-facing summaries must derive from receipts, not from memory.

Forwarded commands (allowlist): `render, validate, deliver, preview, compare, check, inspect, migrate, guide, examples, doctor, demo, visual-check`. `brands` is deliberately NOT forwarded (brand marks include NC-licensed icons).

## Five diagram types

`architecture` (components/boundaries/connections) · `workflow` (lanes/nodes/edges) · `sequence` (participants/messages) · `dataflow` (stages/nodes/flows) · `lifecycle` (lanes/states/transitions). Every IR needs `schema_version`, `diagram_type`, and `meta.title` plus the type-specific arrays. Minimal authoring loop and IR sketch: [references/authoring-cheatsheet.md](references/authoring-cheatsheet.md).

## Evidence binding — architecture only

Only the `architecture` type supports repo evidence validation (`--repo-root` at a pinned revision). The other four types have NO equivalent evidence guarantee — never claim verification for them. Evidence checks must run where the Git checkout/objects live; reading a copied working tree does not validate a pinned revision, and dirty-working-tree checks must never be presented as revision checks. Source links must distinguish pinned revision from current working-tree content — show path/line honestly rather than faking a clickable link.

## Three verification tiers — report separately, never conflate

1. **validate/deliver receipt** — deterministic structure/geometry/evidence checks. A receipt is not a visual check.
2. **browser visual check** — real browser at a named viewport. A visual check is not human review.
3. **human / image review** — readability, visual quality, information organization.

Missing Chrome/Chromium, image review, or a specific environment is reported honestly; a server without a browser still allows render/deliver plus client-side viewing — stating WHERE each check ran. On generation failure the last-good artifact stays visible WITH the failure reason; never present last-good as verified against current inputs.

## Context budget

Only small IR summaries, component IDs, source references, and knowledge pointers enter agent context. Never paste full HTML, embedded assets, or a diagram's DOM into a prompt.

From imBlanker/multi-agents-workflow, MIT.

More from imBlanker/multi-agents-workflow

Signals

Stars
1
Forks
0
Open issues
2
Last commit (this folder)
Sep 17, 2026 (1 month ago)
Last release
Sep 17, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from imBlanker

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get mawf-archify running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying mawf-archify?

CallGet help
Cite this page

mawf-archify skill. Agent Mart. https://agentmart.dev/skills/imblanker-mawf-archify (last updated 2026-10-24)