Deep codebase analysis with progressive disclosure docs in .claude/docs/
/plugin marketplace add settlemint/agent-marketplace/plugin install flow@settlemintThis skill inherits all available tools. When active, it can use any tool Claude has access to.
templates/doc-template.mdtemplates/index-template.mdworkflows/incremental-update.mdworkflows/init-monorepo.mdworkflows/init-standard.mdGenerate modular, discoverable documentation in .claude/docs/ using parallel exploration agents. Creates focused docs (300-600 tokens each) with progressive disclosure patterns. Supports monorepos with per-package documentation.
<quick_start>
// Full codebase analysis (first run or major refresh)
Skill({ skill: "flow:init-enhanced" });
// Incremental update (only changed areas)
Skill({ skill: "flow:init-enhanced", args: "--incremental" });
</quick_start>
<routing>| Pattern | Workflow |
|---|---|
--incremental OR .meta.json is recent | incremental-update |
turbo.json OR pnpm-workspace.yaml found | init-monorepo |
| Default | init-standard |
<output_structure>
Generated docs in .claude/docs/:
.claude/docs/
├── _index.md # Quick reference (~200 tokens)
├── architecture.md # Structure + core logic
├── getting-started.md # Setup + configuration
├── data-layer.md # Database, models, schemas
├── api-reference.md # Routes, endpoints, interfaces
├── testing.md # Test patterns, coverage
├── deployment.md # CI/CD, docker, infra
├── dependencies.md # Key packages explained
├── glossary.md # Domain terminology
├── .meta.json # Staleness tracking
└── packages/ # Per-package docs (monorepo)
├── api/
└── web/
</output_structure>
<agent_strategy>
Launch 5-6 Explore agents in parallel for comprehensive analysis:
| Agent | Focus | Output Doc |
|---|---|---|
| 1 | Dirs, entry points, patterns | architecture.md |
| 2 | Env, build, package manager | getting-started.md |
| 3 | Models, services, domain | data-layer.md + glossary.md |
| 4 | APIs, routes, auth | api-reference.md |
| 5 | Tests, CI, deploy | testing.md + deployment.md |
| 6 | Key packages, internal deps | dependencies.md |
</agent_strategy>
<merge_markers>
Preserve manual edits via markers:
<!-- AUTO-GENERATED: architecture -->
[Generated content]
<!-- END AUTO-GENERATED -->
## Manual Notes
[User additions preserved here]
</merge_markers>
<meta_json_schema>
{
"generated": "2026-01-15T12:00:00Z",
"gitSha": "abc123",
"docs": {
"architecture.md": {
"updated": "2026-01-15T12:00:00Z",
"watchPatterns": ["**/package.json", "turbo.json", "src/**/index.ts"]
}
}
}
</meta_json_schema>
<success_criteria>
.claude/docs/ created with modular files_index.md provides quick reference.meta.json tracks git SHA and watch patterns</success_criteria>
<constraints>.meta.json required - enables incremental updates<anti_patterns>
</anti_patterns>
<related_skills>
| Skill | Relationship |
|---|---|
Skill({ skill: "flow:enhance:explore" }) | Powers the parallel exploration |
Skill({ skill: "devtools:turbo" }) | Monorepo detection patterns |
Skill({ skill: "devtools:drizzle" }) | Data layer documentation |
</related_skills>
<evolution>Extension points:
templates/ directory.meta.jsonCreating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.