tree-generator

Name: tree-generator
Author: TanStack

314stars19forks

How to install tree-generator?

tanstack/intent/tree-generator

$npx -y skills add TanStack/intent --skill tree-generator

Installs into the current project.

Summary

Generate, update, and version a complete skill tree (collection of SKILL.md files) for any JavaScript or TypeScript library. Produces core skills (framework-agnostic) and framework skills (React, Solid, Vue bindings) with dependency linking. Activate when producing skill files fr

Files · 1

View on GitHub

SKILL.md

1	# Skill Tree Generator
2
3	You produce and maintain a tree of SKILL.md files for a library. Every file
4	you create is read directly by AI coding agents across Claude, GPT-4+,
5	Gemini, Cursor, Copilot, Codex, and open-source models. Your output must
6	be portable, concise, and grounded in actual library behavior.
7
8	### Skill types
9
10	Every skill has a `type` field in its frontmatter. Valid types:
11
12	\| Type \| Purpose \| Example \|
13	\| ------------- \| ---------------------------------------------------------- \| ------------------------- \|
14	\| `core` \| Framework-agnostic concepts, configuration, patterns \| `db-core` \|
15	\| `sub-skill` \| A focused sub-topic within a core or framework skill \| `db-core/live-queries` \|
16	\| `framework` \| Framework-specific bindings, hooks, components \| `react-db` \|
17	\| `lifecycle` \| Cross-cutting developer journey (getting started, go-live) \| `electric-quickstart` \|
18	\| `composition` \| Integration between two or more libraries \| `electric-drizzle` \|
19	\| `security` \| Audit checklist or security validation \| `electric-security-check` \|
20
21	Agents discover skills via `npx @tanstack/intent list` and read them directly
22	from `node_modules`. Framework skills declare a `requires` dependency on
23	their core skill so agents load them in the right order.
24
25	There are two workflows. Detect which applies.
26
27	Workflow A — Generate: Build a complete skill tree from a domain map.
28	Workflow B — Update: Diff a library version change and update skills.
29
30	---
31
32	## Workflow A — Generate skill tree
33
34	### Prerequisites
35
36	You need one of:
37
38	- `skills/_artifacts/domain_map.yaml` and `skills/_artifacts/skill_spec.md`
39	from skill-domain-discovery
40	- Raw library documentation and source code (run a compressed domain
41	discovery first)
42
43	If starting from raw docs without a domain map, run a compressed
44	discovery. This produces lower-fidelity output than the full
45	skill-domain-discovery skill — prefer running that when time permits.
46
47	1. Build a concept inventory (every export, config key, constraint, warning)
48	2. Group into capability domains using work-oriented names (let library complexity drive the count — 2–3 for focused libraries, more for large frameworks)
49	3. Enumerate 10–20 task-focused skills from the intersection of domains
50	and developer tasks
51	4. Extract 3+ failure modes per skill (plausible, silent, grounded)
52	5. Proceed to Step 1 below
53
54	### Scaffold flow output
55
56	If the maintainer uses a custom skills root, replace `skills/` in the paths
57	below with their chosen directory.
58
59	For the scaffold workflow, produce a single artifact before writing any
60	SKILL.md files:
61
62	- `skills/_artifacts/skill_tree.yaml`
63
64	This file enumerates every skill that must be generated in the next step.
65	Do not write SKILL.md files yet unless explicitly asked.
66
67	Use this format:
68
69	```yaml
70	# skills/_artifacts/skill_tree.yaml
71	library:
72	name: '[package-name]'
73	version: '[version]'
74	repository: '[repo URL]'
75	description: '[one line]'
76	generated_from:
77	domain_map: 'skills/_artifacts/domain_map.yaml'
78	skill_spec: 'skills/_artifacts/skill_spec.md'
79	generated_at: '[ISO date]'
80
81	skills:
82	- name: '[task-focused skill name]'
83	slug: '[kebab-case]'
84	type: 'core \| sub-skill \| framework \| lifecycle \| composition \| security'
85	domain: '[domain slug]'
86	path: 'skills/[path]/SKILL.md'
87	package: '[package directory, e.g. packages/client]' # monorepo only — which package this skill belongs to
88	description: '[1–2 sentence agent-facing routing key]'
89	requires:
90	- '[other skill slugs]' # omit if none
91	sources:
92	- '[Owner/repo]:docs/[path].md'
93	- '[Owner/repo]:src/[path].ts'
94	subsystems:
95	- '[adapter/backend name]' # omit if none
96	references:
97	- 'references/[file].md' # omit if none
98	```
99
100	Monorepo layout: For monorepos, each skill's `path` is relative to its
101	package directory (e.g. `packages/client/skills/core/SKILL.md`). Set the
102	`package` field so generate-skill knows where to write the file. The domain
103	map artifacts stay at the repo root.
104
105	### Minimal library fast path
106
107	If t

People who liked this also viewed

6 picks

Type

tree-generator

314stars19forks

View on GitHub

How to install tree-generator?

tanstack/intent/tree-generator

$npx -y skills add TanStack/intent --skill tree-generator

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Skill Tree Generator
2
3	You produce and maintain a tree of SKILL.md files for a library. Every file
4	you create is read directly by AI coding agents across Claude, GPT-4+,
5	Gemini, Cursor, Copilot, Codex, and open-source models. Your output must
6	be portable, concise, and grounded in actual library behavior.
7
8	### Skill types
9
10	Every skill has a `type` field in its frontmatter. Valid types:
11
12	\| Type \| Purpose \| Example \|
13	\| ------------- \| ---------------------------------------------------------- \| ------------------------- \|
14	\| `core` \| Framework-agnostic concepts, configuration, patterns \| `db-core` \|
15	\| `sub-skill` \| A focused sub-topic within a core or framework skill \| `db-core/live-queries` \|
16	\| `framework` \| Framework-specific bindings, hooks, components \| `react-db` \|
17	\| `lifecycle` \| Cross-cutting developer journey (getting started, go-live) \| `electric-quickstart` \|
18	\| `composition` \| Integration between two or more libraries \| `electric-drizzle` \|
19	\| `security` \| Audit checklist or security validation \| `electric-security-check` \|
20
21	Agents discover skills via `npx @tanstack/intent list` and read them directly
22	from `node_modules`. Framework skills declare a `requires` dependency on
23	their core skill so agents load them in the right order.
24
25	There are two workflows. Detect which applies.
26
27	Workflow A — Generate: Build a complete skill tree from a domain map.
28	Workflow B — Update: Diff a library version change and update skills.
29
30	---
31
32	## Workflow A — Generate skill tree
33
34	### Prerequisites
35
36	You need one of:
37
38	- `skills/_artifacts/domain_map.yaml` and `skills/_artifacts/skill_spec.md`
39	from skill-domain-discovery
40	- Raw library documentation and source code (run a compressed domain
41	discovery first)
42
43	If starting from raw docs without a domain map, run a compressed
44	discovery. This produces lower-fidelity output than the full
45	skill-domain-discovery skill — prefer running that when time permits.
46
47	1. Build a concept inventory (every export, config key, constraint, warning)
48	2. Group into capability domains using work-oriented names (let library complexity drive the count — 2–3 for focused libraries, more for large frameworks)
49	3. Enumerate 10–20 task-focused skills from the intersection of domains
50	and developer tasks
51	4. Extract 3+ failure modes per skill (plausible, silent, grounded)
52	5. Proceed to Step 1 below
53
54	### Scaffold flow output
55
56	If the maintainer uses a custom skills root, replace `skills/` in the paths
57	below with their chosen directory.
58
59	For the scaffold workflow, produce a single artifact before writing any
60	SKILL.md files:
61
62	- `skills/_artifacts/skill_tree.yaml`
63
64	This file enumerates every skill that must be generated in the next step.
65	Do not write SKILL.md files yet unless explicitly asked.
66
67	Use this format:
68
69	```yaml
70	# skills/_artifacts/skill_tree.yaml
71	library:
72	name: '[package-name]'
73	version: '[version]'
74	repository: '[repo URL]'
75	description: '[one line]'
76	generated_from:
77	domain_map: 'skills/_artifacts/domain_map.yaml'
78	skill_spec: 'skills/_artifacts/skill_spec.md'
79	generated_at: '[ISO date]'
80
81	skills:
82	- name: '[task-focused skill name]'
83	slug: '[kebab-case]'
84	type: 'core \| sub-skill \| framework \| lifecycle \| composition \| security'
85	domain: '[domain slug]'
86	path: 'skills/[path]/SKILL.md'
87	package: '[package directory, e.g. packages/client]' # monorepo only — which package this skill belongs to
88	description: '[1–2 sentence agent-facing routing key]'
89	requires:
90	- '[other skill slugs]' # omit if none
91	sources:
92	- '[Owner/repo]:docs/[path].md'
93	- '[Owner/repo]:src/[path].ts'
94	subsystems:
95	- '[adapter/backend name]' # omit if none
96	references:
97	- 'references/[file].md' # omit if none
98	```
99
100	Monorepo layout: For monorepos, each skill's `path` is relative to its
101	package directory (e.g. `packages/client/skills/core/SKILL.md`). Set the
102	`package` field so generate-skill knows where to write the file. The domain
103	map artifacts stay at the repo root.
104
105	### Minimal library fast path
106
107	If t

People who liked this also viewed

6 picks

Type