repolens-ai 2.0.2 → 3.1.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +127 -4
- package/dist/cli.js +19362 -3643
- package/package.json +21 -10
- package/dist/cli.js.map +0 -1
package/README.md
CHANGED
|
@@ -30,11 +30,33 @@ The problem isn't missing context files — it's **bad** context files.
|
|
|
30
30
|
RepoLens AI is a **context quality platform** that helps you:
|
|
31
31
|
|
|
32
32
|
- ⭐ **Setup** AI context for 6 tools with ONE command
|
|
33
|
+
- **Context** — create a paste-ready project brief for ChatGPT, Claude, Gemini, Cursor, and Antigravity
|
|
34
|
+
- **Prompt** — turn vague ideas into project-aware AI coding prompts
|
|
35
|
+
- **Check** — sanity-check AI-generated changes before commit
|
|
33
36
|
- 📊 **Score** your context files quality (0-100, 5 dimensions)
|
|
34
37
|
- 🔄 **Sync** AGENTS.md → Cursor (.mdc) → Claude → Copilot → Windsurf → Codex
|
|
35
38
|
- 📚 **Templates** — 12 curated framework templates with real gotchas
|
|
36
39
|
- 🔧 **Fix** — auto-remove generic rules that hurt AI performance
|
|
37
40
|
|
|
41
|
+
## For Antigravity & Vibe Coding
|
|
42
|
+
|
|
43
|
+
RepoLens does **not** replace Antigravity, Cursor, Claude Code, Codex, or
|
|
44
|
+
Copilot. It prepares your repository so those agents stop guessing.
|
|
45
|
+
|
|
46
|
+
```bash
|
|
47
|
+
npx repolens-ai setup
|
|
48
|
+
npx repolens-ai context
|
|
49
|
+
npx repolens-ai prompt "add login with email verification"
|
|
50
|
+
npx repolens-ai check
|
|
51
|
+
npx repolens-ai vibe
|
|
52
|
+
```
|
|
53
|
+
|
|
54
|
+
`setup` creates high-signal project context. `context` gives any AI chat the
|
|
55
|
+
short brief it needs. `prompt` turns an idea into a repo-aware instruction.
|
|
56
|
+
`check` catches obvious AI-generated mistakes before they become project debt.
|
|
57
|
+
`vibe` tells you whether the repo is ready for AI-assisted work and what to
|
|
58
|
+
improve next.
|
|
59
|
+
|
|
38
60
|
### Supported AI Tools
|
|
39
61
|
|
|
40
62
|
| Tool | File Generated | Format |
|
|
@@ -63,6 +85,9 @@ cd my-project
|
|
|
63
85
|
|
|
64
86
|
# ONE COMMAND — sets up everything
|
|
65
87
|
repolens setup
|
|
88
|
+
|
|
89
|
+
# Optional: see whether the repo is ready for AI-assisted work
|
|
90
|
+
repolens vibe
|
|
66
91
|
```
|
|
67
92
|
|
|
68
93
|
That's it. `repolens setup` will:
|
|
@@ -72,10 +97,17 @@ That's it. `repolens setup` will:
|
|
|
72
97
|
4. ✅ Generate skill files for Cursor
|
|
73
98
|
5. ✅ Score your context quality
|
|
74
99
|
|
|
100
|
+
`repolens vibe` then gives you one community-readiness score across context
|
|
101
|
+
quality, AI tool coverage, workflow clarity, and code health.
|
|
102
|
+
|
|
75
103
|
**Want more control?** Use individual commands:
|
|
76
104
|
|
|
77
105
|
```bash
|
|
106
|
+
repolens context # Paste-ready context for any AI chat
|
|
107
|
+
repolens prompt "..." # Project-aware prompt generator
|
|
108
|
+
repolens check # Sanity-check AI-generated changes
|
|
78
109
|
repolens init # Interactive AGENTS.md creation (10 questions)
|
|
110
|
+
repolens vibe # Score vibe-coding readiness
|
|
79
111
|
repolens lint # Score your context quality (0-100)
|
|
80
112
|
repolens fix # Auto-fix generic rules
|
|
81
113
|
repolens sync # Sync to all AI tools
|
|
@@ -85,7 +117,67 @@ repolens dashboard # Visual web dashboard
|
|
|
85
117
|
|
|
86
118
|
## Commands
|
|
87
119
|
|
|
88
|
-
###
|
|
120
|
+
### Daily Vibe Coding Loop
|
|
121
|
+
|
|
122
|
+
Use these commands every day with Cursor, Antigravity, Claude, ChatGPT, Gemini,
|
|
123
|
+
Codex, or any AI coding workflow.
|
|
124
|
+
|
|
125
|
+
```bash
|
|
126
|
+
repolens context
|
|
127
|
+
repolens prompt "add image upload to user profiles"
|
|
128
|
+
repolens check
|
|
129
|
+
```
|
|
130
|
+
|
|
131
|
+
#### `repolens context`
|
|
132
|
+
|
|
133
|
+
Creates a compact project brief that can be pasted into any AI chat. It includes
|
|
134
|
+
the detected stack, architecture, key directories, critical files, rules,
|
|
135
|
+
commands, and gotchas.
|
|
136
|
+
|
|
137
|
+
```bash
|
|
138
|
+
repolens context
|
|
139
|
+
repolens context --copy
|
|
140
|
+
```
|
|
141
|
+
|
|
142
|
+
#### `repolens prompt "<request>"`
|
|
143
|
+
|
|
144
|
+
Turns a vague feature idea into a project-aware prompt. RepoLens detects the
|
|
145
|
+
task intent, finds the most relevant project files, suggests likely files to
|
|
146
|
+
create or update, adds feature-specific guardrails, and gives the AI agent a
|
|
147
|
+
clear verification plan.
|
|
148
|
+
|
|
149
|
+
```bash
|
|
150
|
+
repolens prompt "add customer avatar upload"
|
|
151
|
+
repolens prompt "add customer avatar upload" --copy
|
|
152
|
+
```
|
|
153
|
+
|
|
154
|
+
#### `repolens check`
|
|
155
|
+
|
|
156
|
+
Fast guardrail after AI edits. It scans changed files for risky changes,
|
|
157
|
+
TypeScript `any`, missing `use client` in React hook files, and debug logging.
|
|
158
|
+
|
|
159
|
+
```bash
|
|
160
|
+
repolens check
|
|
161
|
+
```
|
|
162
|
+
|
|
163
|
+
### 🆕 Context Intelligence
|
|
164
|
+
|
|
165
|
+
#### `repolens vibe`
|
|
166
|
+
|
|
167
|
+
Score whether a repository is ready for vibe coding and community use. This is
|
|
168
|
+
the fastest way to see if a repo has enough context, tool coverage, workflow
|
|
169
|
+
clarity, and code-health signals for AI agents to work effectively.
|
|
170
|
+
|
|
171
|
+
```bash
|
|
172
|
+
repolens vibe
|
|
173
|
+
```
|
|
174
|
+
|
|
175
|
+
**Checks:**
|
|
176
|
+
- Context quality and drift
|
|
177
|
+
- Coverage for AGENTS.md, Claude/Codex, Cursor, Copilot, Windsurf, and skills
|
|
178
|
+
- Dev/build/lint/test scripts
|
|
179
|
+
- README, CI, license, and contribution readiness
|
|
180
|
+
- Security risks, circular dependencies, complexity, and documentation signals
|
|
89
181
|
|
|
90
182
|
#### `repolens init`
|
|
91
183
|
|
|
@@ -217,6 +309,9 @@ You should! But RepoLens AI helps you write **better** ones:
|
|
|
217
309
|
| You don't know if context is outdated | Freshness scoring detects stale references |
|
|
218
310
|
| You can't visualize context quality | Dashboard shows scores and issues visually |
|
|
219
311
|
|
|
312
|
+
See a concrete [before/after example](examples/before-after-agents.md) of weak
|
|
313
|
+
generic context versus context that actually helps AI agents.
|
|
314
|
+
|
|
220
315
|
## Privacy First
|
|
221
316
|
|
|
222
317
|
- ✅ **Never uploads code** — everything runs locally
|
|
@@ -229,6 +324,28 @@ You should! But RepoLens AI helps you write **better** ones:
|
|
|
229
324
|
|
|
230
325
|
Auto-detects 15+ frameworks: Next.js, React, Vue, Angular, SvelteKit, NestJS, Express, Fastify, Hono, Laravel, Django, FastAPI, Flask, Rails, Go, Rust, Odoo, and more.
|
|
231
326
|
|
|
327
|
+
## Project Structure
|
|
328
|
+
|
|
329
|
+
RepoLens keeps the source tree intentionally simple for contributors and AI
|
|
330
|
+
coding agents:
|
|
331
|
+
|
|
332
|
+
```text
|
|
333
|
+
src/
|
|
334
|
+
cli.ts CLI entrypoint and command registration
|
|
335
|
+
commands/ One file per user-facing command
|
|
336
|
+
core/ Analysis, scoring, sync, and workflow engines
|
|
337
|
+
ai/ Optional AI enhancement and privacy filtering
|
|
338
|
+
reporters/ Terminal and markdown output
|
|
339
|
+
templates/ Framework and tool context templates
|
|
340
|
+
utils/ Shared logging, paths, masking, and error handling
|
|
341
|
+
tests/ Focused Vitest coverage for core behavior
|
|
342
|
+
docs/ Command docs, privacy notes, roadmap
|
|
343
|
+
examples/ Before/after and sample context files
|
|
344
|
+
```
|
|
345
|
+
|
|
346
|
+
Build output lives in `dist/` and is intentionally ignored in git. The npm
|
|
347
|
+
package publishes the generated `dist` bundle after `npm run build`.
|
|
348
|
+
|
|
232
349
|
## AI Enhancement (Optional)
|
|
233
350
|
|
|
234
351
|
Works **perfectly without any API key**. For AI-enhanced explanations:
|
|
@@ -240,7 +357,9 @@ repolens analyze
|
|
|
240
357
|
|
|
241
358
|
## Roadmap
|
|
242
359
|
|
|
243
|
-
- [x]
|
|
360
|
+
- [x] Vibe readiness scoring command
|
|
361
|
+
- [x] Daily vibe coding commands: context, prompt, check
|
|
362
|
+
- [x] 18 CLI commands
|
|
244
363
|
- [x] Interactive knowledge extraction
|
|
245
364
|
- [x] Context quality scoring (5 dimensions)
|
|
246
365
|
- [x] Cross-tool sync (AGENTS.md → CLAUDE.md → .cursorrules → Copilot)
|
|
@@ -257,15 +376,19 @@ repolens analyze
|
|
|
257
376
|
## Contributing
|
|
258
377
|
|
|
259
378
|
```bash
|
|
260
|
-
git clone https://github.com/
|
|
379
|
+
git clone https://github.com/Harry-Kien/repolens-ai.git
|
|
261
380
|
cd repolens-ai
|
|
262
381
|
npm install
|
|
382
|
+
npm run typecheck
|
|
383
|
+
npm test
|
|
263
384
|
npm run build
|
|
264
385
|
npm link
|
|
265
386
|
|
|
266
387
|
# Test locally
|
|
388
|
+
repolens context
|
|
389
|
+
repolens prompt "add login"
|
|
390
|
+
repolens check
|
|
267
391
|
repolens doctor
|
|
268
|
-
repolens lint
|
|
269
392
|
```
|
|
270
393
|
|
|
271
394
|
## License
|