h4tcher-docusaurus-generator

Name: h4tcher-docusaurus-generator
Author: hatch3r

27stars4forks

How to install h4tcher-docusaurus-generator?

hatch3r/hatch3r/h4tcher-docusaurus-generator

$npx -y skills add hatch3r/hatch3r --skill h4tcher-docusaurus-generator

Installs into the current project.

Summary

Maintainer skill — generate or refresh the hatch3r project documentation site (Docusaurus 3.x) from this repo's canonical corpus (agents/, skills/, rules/, commands/, hooks/, docs/, plus counts from the public governance/inventory.json). Use when a hatch3r maintainer asks to buil

Files · 1

View on GitHub

SKILL.md

1	> Last updated: 2026-07-06
2
3	# Docusaurus Generator (hatch3r maintainer skill)
4
5	This skill refreshes the hatch3r project's own documentation site, which already exists at `website/` (Docusaurus 3.x), by re-deriving its pages from this repository's canonical corpus. It is a framework-dev maintainer tool (hence the `h4tcher-` prefix per CLAUDE.md), not a generic per-project docs generator and not a lifecycle add/refactor/remove preset. It sources content from `agents/`, `skills/`, `rules/`, `commands/`, `hooks/`, and `docs/`; the only governance source is the public `governance/inventory.json` (counts).
6
7	The site lives in `website/` and deploys to `docs.hatch3r.com` via `.github/workflows/deploy-docs.yml` on push to `main` (paths filter: `website/**`). There is no `docs-site/` directory — operate on `website/` directly. Do NOT run `npx create-docusaurus`: the project is already scaffolded; creating a second project is a P4 (Lean Coverage) violation.
8
9	## Workflow Overview
10
11	1. Analyze the hatch3r repo corpus to identify what each page must reflect
12	2. Map corpus sources to the existing `website/docs/**` page set and `website/sidebars.ts`
13	3. Regenerate the affected pages in place under `website/docs/`
14	4. Build & Preview via the website's own npm scripts
15
16	## Step 1: Analyze the hatch3r corpus
17
18	Scan this repo to identify what each page must reflect:
19
20	- CLI surface: the commands under `src/cli/commands/` (mirror the `npx hatch3r <cmd>` help) → `website/docs/reference/commands/cli-commands.md`
21	- Canonical content: `agents/`, `skills/`, `rules/`, `commands/`, `hooks/` — the artifacts hatch3r ships → the matching `website/docs/reference/{agents,skills,rules,hooks}.md` pages and `reference/commands/agent-commands.md`
22	- Governance: summarize public-facing governance behavior only, in your own prose, on architecture and guide pages. The sole citable governance source is the public `governance/inventory.json` (counts). NEVER emit private governance file paths, § numbers, or Decision numbers into public pages — the private governance corpus may exist locally via the overlay, but it is not a citable source for the public site
23	- Existing docs: the `docs/` directory and `README.md` (do not duplicate; link or lift)
24	- Source-of-truth counts: `governance/inventory.json` `counts` — page tallies (agents, skills, rules, commands, hooks) must match it
25
26	```bash
27	# Key sources to examine in the hatch3r repo
28	ls src/cli/commands/*.ts
29	ls agents/ skills/ rules/ commands/ hooks/
30	ls docs/
31	jq '.name, .description' package.json
32	jq '.counts' governance/inventory.json
33	```
34
35	## Step 2: Map corpus to the existing site
36
37	The site's page set and navigation are already defined. Read both before editing so regeneration lands in the right files instead of creating new ones:
38
39	```bash
40	# Existing page tree and navigation — read before regenerating
41	find website/docs -name '*.md' \| sort
42	cat website/sidebars.ts # docsSidebar: category -> doc-id ordering
43	cat website/docusaurus.config.ts # routeBasePath 'docs', navbar/footer links, mermaid theme
44	```
45
46	Current top-level groups in `website/sidebars.ts` (`docsSidebar`): `about`, Getting Started, Guides, Reference (with Architecture and Commands sub-categories), `troubleshooting`. Every doc id listed in `sidebars.ts` MUST resolve to a file under `website/docs/` — `onBrokenLinks: 'throw'` and the sidebar-id check fail the build otherwise.
47
48	Mapping rule: a corpus change updates the existing page that already documents that surface. Add a new page only when a new artifact class or CLI command has no home; when you add one, add its doc id to the matching category in `website/sidebars.ts` in the same change.
49
50	## Step 3: Regenerate pages in place
51
52	Rewrite the affected files under `website/docs/**`. Keep each page's existing frontmatter shape:
53
54	```markdown
55	---
56	sidebar_position: 1
57	title: Page Title
58	description: Brief description for SEO
59	---
60
61	# Page Title
62
63	Content here...
64	```
65
66	Content guidelines:
67
68	- Write for hatch3r users (people running `npx hatch3r`), not framework contributors.
69	- Keep counts and artifact lists synchronized with `governance/inventory.json` `counts`.
70	- Link between related docs with relative doc paths (e.g. `[workflow](../guides/workflow)`), not absolute URLs, so the broken-link check covers them.
71	- Do not duplicate `README.md` or `docs/` prose — link to it or lift the canonical wording

People who liked this also viewed

6 picks

Type

h4tcher-docusaurus-generator

27stars4forks

View on GitHub

How to install h4tcher-docusaurus-generator?

hatch3r/hatch3r/h4tcher-docusaurus-generator

$npx -y skills add hatch3r/hatch3r --skill h4tcher-docusaurus-generator

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	> Last updated: 2026-07-06
2
3	# Docusaurus Generator (hatch3r maintainer skill)
4
5	This skill refreshes the hatch3r project's own documentation site, which already exists at `website/` (Docusaurus 3.x), by re-deriving its pages from this repository's canonical corpus. It is a framework-dev maintainer tool (hence the `h4tcher-` prefix per CLAUDE.md), not a generic per-project docs generator and not a lifecycle add/refactor/remove preset. It sources content from `agents/`, `skills/`, `rules/`, `commands/`, `hooks/`, and `docs/`; the only governance source is the public `governance/inventory.json` (counts).
6
7	The site lives in `website/` and deploys to `docs.hatch3r.com` via `.github/workflows/deploy-docs.yml` on push to `main` (paths filter: `website/**`). There is no `docs-site/` directory — operate on `website/` directly. Do NOT run `npx create-docusaurus`: the project is already scaffolded; creating a second project is a P4 (Lean Coverage) violation.
8
9	## Workflow Overview
10
11	1. Analyze the hatch3r repo corpus to identify what each page must reflect
12	2. Map corpus sources to the existing `website/docs/**` page set and `website/sidebars.ts`
13	3. Regenerate the affected pages in place under `website/docs/`
14	4. Build & Preview via the website's own npm scripts
15
16	## Step 1: Analyze the hatch3r corpus
17
18	Scan this repo to identify what each page must reflect:
19
20	- CLI surface: the commands under `src/cli/commands/` (mirror the `npx hatch3r <cmd>` help) → `website/docs/reference/commands/cli-commands.md`
21	- Canonical content: `agents/`, `skills/`, `rules/`, `commands/`, `hooks/` — the artifacts hatch3r ships → the matching `website/docs/reference/{agents,skills,rules,hooks}.md` pages and `reference/commands/agent-commands.md`
22	- Governance: summarize public-facing governance behavior only, in your own prose, on architecture and guide pages. The sole citable governance source is the public `governance/inventory.json` (counts). NEVER emit private governance file paths, § numbers, or Decision numbers into public pages — the private governance corpus may exist locally via the overlay, but it is not a citable source for the public site
23	- Existing docs: the `docs/` directory and `README.md` (do not duplicate; link or lift)
24	- Source-of-truth counts: `governance/inventory.json` `counts` — page tallies (agents, skills, rules, commands, hooks) must match it
25
26	```bash
27	# Key sources to examine in the hatch3r repo
28	ls src/cli/commands/*.ts
29	ls agents/ skills/ rules/ commands/ hooks/
30	ls docs/
31	jq '.name, .description' package.json
32	jq '.counts' governance/inventory.json
33	```
34
35	## Step 2: Map corpus to the existing site
36
37	The site's page set and navigation are already defined. Read both before editing so regeneration lands in the right files instead of creating new ones:
38
39	```bash
40	# Existing page tree and navigation — read before regenerating
41	find website/docs -name '*.md' \| sort
42	cat website/sidebars.ts # docsSidebar: category -> doc-id ordering
43	cat website/docusaurus.config.ts # routeBasePath 'docs', navbar/footer links, mermaid theme
44	```
45
46	Current top-level groups in `website/sidebars.ts` (`docsSidebar`): `about`, Getting Started, Guides, Reference (with Architecture and Commands sub-categories), `troubleshooting`. Every doc id listed in `sidebars.ts` MUST resolve to a file under `website/docs/` — `onBrokenLinks: 'throw'` and the sidebar-id check fail the build otherwise.
47
48	Mapping rule: a corpus change updates the existing page that already documents that surface. Add a new page only when a new artifact class or CLI command has no home; when you add one, add its doc id to the matching category in `website/sidebars.ts` in the same change.
49
50	## Step 3: Regenerate pages in place
51
52	Rewrite the affected files under `website/docs/**`. Keep each page's existing frontmatter shape:
53
54	```markdown
55	---
56	sidebar_position: 1
57	title: Page Title
58	description: Brief description for SEO
59	---
60
61	# Page Title
62
63	Content here...
64	```
65
66	Content guidelines:
67
68	- Write for hatch3r users (people running `npx hatch3r`), not framework contributors.
69	- Keep counts and artifact lists synchronized with `governance/inventory.json` `counts`.
70	- Link between related docs with relative doc paths (e.g. `[workflow](../guides/workflow)`), not absolute URLs, so the broken-link check covers them.
71	- Do not duplicate `README.md` or `docs/` prose — link to it or lift the canonical wording

People who liked this also viewed

6 picks

Type