sickn33/agentic-awesome-skills / plugins/agentic-awesome-skills-claude/skills/agents-generator
agents-generator
VerifiedGenerate project-specific AGENTS.md and companion rules by analyzing a codebase. Supports full, minimal, update, and dry-run modes with package-manager detection, monorepos, backups, managed blocks, confidence scoring, and command validation.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 9.7 KB |
| assets/agents-full.md | markdown | 11.8 KB |
| assets/agents-minimal.md | markdown | 1.8 KB |
| assets/agents-nested.md | markdown | 1.8 KB |
| assets/architecture.md | markdown | 1.6 KB |
| assets/backend.md | markdown | 1.5 KB |
| assets/claude.md | markdown | 1.1 KB |
| assets/database.md | markdown | 1.4 KB |
| assets/forms.md | markdown | 1.7 KB |
| assets/frontend-patterns.md | markdown | 1.3 KB |
| assets/git-workflow.md | markdown | 1.8 KB |
| assets/i18n.md | markdown | 1.5 KB |
| assets/platform.md | markdown | 2.8 KB |
| assets/sdd-workflow.md | markdown | 1.7 KB |
| assets/server-actions.md | markdown | 1.0 KB |
| assets/styling.md | markdown | 1.9 KB |
| assets/testing.md | markdown | 1.9 KB |
| references/decision-matrix.md | markdown | 18.5 KB |
| references/example-output/README.md | markdown | 6.0 KB |
| references/template-filling-guide.md | markdown | 5.3 KB |
references/template-filling-guide.md
View on GitHub# Template Filling Guide
When filling template placeholders in `assets/*.md`, follow these concreteness rules. The output must be as specific as the example in `references/example-output/README.md`.
## Quick Decision Table
Full detection logic is in `decision-matrix.md`. This is the quick-reference table for selecting templates:
| Signal | Decision |
| ------------------------------------- | ------------------------------------------------------------------------------------------------- |
| `workspaces` in package.json | Monorepo → generate package dependency rules, shared package rebuild commands |
| `turbo.json`, `nx.json`, `lerna.json` | Monorepo tool → adapt commands (`turbo run`, `nx`, `lerna run`) |
| `next` in deps + `app/` vs `pages/` | Next.js Router type → App Router (server components default) vs Pages Router (getServerSideProps) |
| `@nestjs/core` | NestJS backend → module architecture, guards, DTOs |
| `@trpc/*` | tRPC → completely different backend rules than REST |
| `react-hook-form`, `formik` | Form library → generate `forms.md` |
| `next-intl`, `react-i18next` | i18n → generate `i18n.md` |
| `prisma`, `drizzle-orm` | ORM → generate `database.md` |
| `styled-components`, `.module.css` | Non-Tailwind CSS → generate `styling.md` with different rules |
| `next-auth`, `@clerk/nextjs`, `lucia` | Auth → route protection rules |
| `@tanstack/react-query`, `swr` | Server state → different state rules than Zustand/Redux |
| `vitest`, `jest`, `playwright` | Test runner(s) → generate `testing.md` with correct commands |
| No TypeScript | JS project → skip `tsc`, different type safety rules |
## Placeholder Filling Rules
### `{{STACK_TABLE}}`
Report declared ranges from `package.json` as ranges; resolve exact installed versions from the lockfile when available. Do not relabel a semver range as an exact version. Never write "latest" or "X.Y.Z". Format: `| Capa | Tecnología | Versión |`. Include framework, UI, runtime, ORM, validation, testing, linting, analytics, fonts.
### `{{ARCHITECTURE_DIAGRAM}}`
ASCII art using **actual directory names** from the project. Show real connections — which file imports which. No boxes labeled "Component" or "Service". For monorepos, show package dependency arrows. For single apps, show app/ → components/ + lib/ → external services.
### `{{DATA_FLOW}}`
Numbered steps with **real function names**: `detectPlatform()` not "validate input"; `ttdl(url)` not "call API". Show the complete chain: user action → frontend → backend/server action → external library/database → response → UI update. Use ASCII arrows between steps.
### `{{ROUTING_TABLE}}`
**Every route file** found in `app/` or `pages/`. Include: route path, type (server/client/API), file path, and 1-line purpose in the project's UI language. Check for: `page.tsx`, `layout.tsx`, `route.ts`, `error.tsx`, `loading.tsx`, `not-found.tsx`, `sitemap.ts`, `robots.ts`, API routes, dynamic routes.
### `{{ESSENTIAL_COMMANDS}}`
Only commands whose **script key exists** in `package.json`. Use `{pm} run <key>` or `{pm} <key>` format matching the package manager. If a script is missing, omit that row — don't invent commands. Core rows: dev, build, lint, test, test:run, test:coverage, and quality tool (doctor/lint-staged) if present.
### `{{TEST_COUNT}}`
Static `test()` / `it()` call counts miss parameterized/generated cases and are not executed-test totals. Label them static declarations, or report test-file counts. If you cannot count accurately, report the number of test files: `"N test files"`.
Also report: test runner + version, environment (node/jsdom), coverage provider and target, strict TDD mode (yes/no).
### `{{COMMIT_EXAMPLES}}`
2-3 examples using **real file names** or feature names from the project's recent git history. Read `git log --oneline -5` if available to get realistic examples. Format: `type: message`, one per line.
### `{{VERIFICATION_CYCLE}}`
Assemble from validated commands: `{typecheck} → {lint} → {test} → {doctor}`. Only include steps whose tool exists. Use the exact command format from package.json.
### `{{GLOBAL_CONVENTIONS}}`
List 5-8 project-specific conventions. Each must be verifiable from config files, not generic advice. Examples:
- ✅ "Bun siempre. No npm ni yarn." (detected from bun.lock)
- ✅ "Sin librería de validación externa — TypeScript types + guards manuales." (no zod in deps)
- ❌ "Usar buenas prácticas de TypeScript." (generic, unverifiable)
- ❌ "Código limpio y documentado." (generic, unverifiable)
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |