agdi 2.6.0 → 2.6.2
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 +156 -406
- package/dist/index.js +313 -224
- package/package.json +8 -3
package/README.md
CHANGED
|
@@ -1,25 +1,24 @@
|
|
|
1
1
|
<div align="center">
|
|
2
2
|
|
|
3
3
|
```
|
|
4
|
-
|
|
5
|
-
|
|
6
|
-
|
|
7
|
-
|
|
8
|
-
/_/ |_
|
|
9
|
-
|
|
4
|
+
___ __________ ____
|
|
5
|
+
/ | / ____/ __ \/ _/
|
|
6
|
+
/ /| |/ / __/ / / // /
|
|
7
|
+
/ ___ / /_/ / /_/ // /
|
|
8
|
+
/_/ |_\____/_____/___/
|
|
9
|
+
|
|
10
10
|
```
|
|
11
11
|
|
|
12
12
|
# agdi
|
|
13
13
|
|
|
14
|
-
### The AI-Powered
|
|
14
|
+
### The AI-Powered Terminal Architect
|
|
15
15
|
|
|
16
|
-
[](https://www.typescriptlang.org/)
|
|
16
|
+
[](https://www.npmjs.com/package/agdi)
|
|
17
|
+
[](https://www.npmjs.com/package/agdi)
|
|
18
|
+
[](https://opensource.org/licenses/MIT)
|
|
19
|
+
[](https://nodejs.org/)
|
|
21
20
|
|
|
22
|
-
**Build
|
|
21
|
+
**Build apps from natural language • Multi-Agent AI • Cinematic Dark Mode**
|
|
23
22
|
|
|
24
23
|
[Website](https://agdi-dev.vercel.app) · [npm](https://www.npmjs.com/package/agdi) · [GitHub](https://github.com/anassagd432/Agdi-dev)
|
|
25
24
|
|
|
@@ -27,27 +26,83 @@
|
|
|
27
26
|
|
|
28
27
|
---
|
|
29
28
|
|
|
30
|
-
##
|
|
29
|
+
## 🎬 Cinematic Dark Mode (v2.6.1)
|
|
30
|
+
|
|
31
|
+
<div align="center">
|
|
32
|
+
<img src="https://img.shields.io/badge/NEW-Premium_Terminal_UI-blueviolet?style=for-the-badge" alt="New">
|
|
33
|
+
</div>
|
|
34
|
+
|
|
35
|
+
### ✨ Visual Upgrade
|
|
36
|
+
|
|
37
|
+
| Component | What It Looks Like |
|
|
38
|
+
|-----------|-------------------|
|
|
39
|
+
| **Neon Banner** | Cyan-to-Purple gradient ASCII art on startup |
|
|
40
|
+
| **Glass Cards** | Rounded boxen panels for plans and status |
|
|
41
|
+
| **Chat Bubbles** | `👤 YOU ›` and `⚡ AGDI ›` styled prompts |
|
|
42
|
+
| **Security Alerts** | Red double-bordered warnings |
|
|
43
|
+
| **Spinners** | Smooth cyan dots animation |
|
|
44
|
+
|
|
45
|
+
### 🖥️ Startup Experience
|
|
46
|
+
|
|
47
|
+
```
|
|
48
|
+
┌──────────────────────────────────────────────────────────────┐
|
|
49
|
+
│ │
|
|
50
|
+
│ ___ __________ ____ │
|
|
51
|
+
│ / | / ____/ __ \/ _/ ← Neon Gradient │
|
|
52
|
+
│ / /| |/ / __/ / / // / │
|
|
53
|
+
│ / ___ / /_/ / /_/ // / │
|
|
54
|
+
│ /_/ |_\____/_____/___/ │
|
|
55
|
+
│ │
|
|
56
|
+
│ v2.6.1 [ARCHITECT ONLINE] │
|
|
57
|
+
│ │
|
|
58
|
+
├──────────────────────────────────────────────────────────────┤
|
|
59
|
+
│ ╭─ SESSION INFO ─────────────────────────────────────────╮ │
|
|
60
|
+
│ │ │ │
|
|
61
|
+
│ │ Model: gemini-2.5-flash │ │
|
|
62
|
+
│ │ Workspace: ~/projects/my-app │ │
|
|
63
|
+
│ │ Context: Git Repository │ │
|
|
64
|
+
│ │ │ │
|
|
65
|
+
│ ╰────────────────────────────────────────────────────────╯ │
|
|
66
|
+
│ │
|
|
67
|
+
│ ──────────────────────────────────────────────────────────── │
|
|
68
|
+
│ │
|
|
69
|
+
│ 👤 YOU › create a dashboard with charts │
|
|
70
|
+
│ │
|
|
71
|
+
│ ⚡ AGDI › │
|
|
72
|
+
│ I'll create a modern dashboard with Recharts... │
|
|
73
|
+
│ │
|
|
74
|
+
└──────────────────────────────────────────────────────────────┘
|
|
75
|
+
```
|
|
76
|
+
|
|
77
|
+
---
|
|
78
|
+
|
|
79
|
+
## ⚡ Quick Start
|
|
80
|
+
|
|
81
|
+
```bash
|
|
82
|
+
npm install -g agdi
|
|
83
|
+
agdi
|
|
84
|
+
```
|
|
85
|
+
|
|
86
|
+
---
|
|
87
|
+
|
|
88
|
+
## 🎯 What's New in v2.6
|
|
31
89
|
|
|
32
90
|
<table>
|
|
33
91
|
<tr>
|
|
34
92
|
<td width="50%">
|
|
35
93
|
|
|
36
94
|
### 🤖 Multi-Agent System
|
|
37
|
-
Planner → Coder → Reviewer pipeline. Your code is reviewed by AI before delivery.
|
|
38
|
-
|
|
39
95
|
```bash
|
|
40
96
|
→ /agent
|
|
41
97
|
🤖 Multi-agent mode: ON
|
|
42
|
-
|
|
98
|
+
Planner → Coder → Reviewer
|
|
43
99
|
```
|
|
100
|
+
Your code is reviewed by AI before delivery.
|
|
44
101
|
|
|
45
102
|
</td>
|
|
46
103
|
<td width="50%">
|
|
47
104
|
|
|
48
105
|
### 🔍 Codebase RAG
|
|
49
|
-
Index and search your entire project semantically. The AI understands your code.
|
|
50
|
-
|
|
51
106
|
```bash
|
|
52
107
|
→ /index
|
|
53
108
|
✓ Indexed 47 files, 312 chunks
|
|
@@ -61,9 +116,7 @@ Index and search your entire project semantically. The AI understands your code.
|
|
|
61
116
|
<tr>
|
|
62
117
|
<td width="50%">
|
|
63
118
|
|
|
64
|
-
### 🔧 MCP
|
|
65
|
-
Model Context Protocol for standardized AI tool calls.
|
|
66
|
-
|
|
119
|
+
### 🔧 MCP Tools
|
|
67
120
|
```bash
|
|
68
121
|
→ /tools
|
|
69
122
|
🔧 Available Tools
|
|
@@ -76,8 +129,6 @@ Model Context Protocol for standardized AI tool calls.
|
|
|
76
129
|
<td width="50%">
|
|
77
130
|
|
|
78
131
|
### 🧠 Persistent Memory
|
|
79
|
-
AI remembers your preferences and project context across sessions.
|
|
80
|
-
|
|
81
132
|
```bash
|
|
82
133
|
→ /memory
|
|
83
134
|
🧠 Memory Stats
|
|
@@ -92,433 +143,134 @@ AI remembers your preferences and project context across sessions.
|
|
|
92
143
|
|
|
93
144
|
---
|
|
94
145
|
|
|
95
|
-
##
|
|
96
|
-
|
|
97
|
-
```bash
|
|
98
|
-
# Install globally
|
|
99
|
-
npm install -g agdi
|
|
100
|
-
|
|
101
|
-
# Start coding
|
|
102
|
-
agdi
|
|
103
|
-
```
|
|
104
|
-
|
|
105
|
-
That's it. Agdi handles the rest.
|
|
106
|
-
|
|
107
|
-
---
|
|
108
|
-
|
|
109
|
-
## 🎬 Demo
|
|
110
|
-
|
|
111
|
-
```
|
|
112
|
-
$ agdi
|
|
113
|
-
|
|
114
|
-
⚡ Agdi dev
|
|
115
|
-
|
|
116
|
-
Model: gemini-2.5-flash
|
|
117
|
-
Workspace: ~/projects/my-app
|
|
118
|
-
Commands: /status, /diff, /commit, /build, /index, /search, /agent, /help, /exit
|
|
119
|
-
|
|
120
|
-
──────────────────────────────────────────────────
|
|
121
|
-
|
|
122
|
-
→ /agent
|
|
123
|
-
🤖 Multi-agent mode: ON
|
|
124
|
-
|
|
125
|
-
→ create a todo app with dark mode and local storage
|
|
126
|
-
|
|
127
|
-
📋 Phase 1: Planning...
|
|
128
|
-
✓ Plan created
|
|
129
|
-
|
|
130
|
-
💻 Phase 2: Coding...
|
|
131
|
-
✓ Code generated
|
|
132
|
-
|
|
133
|
-
🔍 Phase 3: Review (iteration 1)...
|
|
134
|
-
✓ Code approved!
|
|
135
|
-
|
|
136
|
-
✅ Multi-Agent Pipeline Complete
|
|
137
|
-
|
|
138
|
-
📋 Action Plan - todo-app
|
|
139
|
-
─────────────────────────
|
|
140
|
-
✓ mkdir src
|
|
141
|
-
✓ writeFile package.json
|
|
142
|
-
✓ writeFile src/App.tsx
|
|
143
|
-
✓ writeFile src/index.css
|
|
144
|
-
✓ exec pnpm install
|
|
145
|
-
|
|
146
|
-
✅ Project created! Run: cd todo-app && pnpm dev
|
|
147
|
-
```
|
|
148
|
-
|
|
149
|
-
---
|
|
150
|
-
|
|
151
|
-
## ✨ Features
|
|
152
|
-
|
|
153
|
-
<table>
|
|
154
|
-
<tr>
|
|
155
|
-
<td width="50%">
|
|
156
|
-
|
|
157
|
-
### 🤖 AI-Powered Development
|
|
158
|
-
- **Natural language to code** — Describe what you want, get production-ready apps
|
|
159
|
-
- **Conversation memory** — AI remembers context across messages
|
|
160
|
-
- **Multi-agent mode** — Planner, Coder, Reviewer pipeline
|
|
161
|
-
- **Codebase RAG** — AI understands your entire project
|
|
162
|
-
|
|
163
|
-
</td>
|
|
164
|
-
<td width="50%">
|
|
165
|
-
|
|
166
|
-
### 🔧 Built-in Git Integration
|
|
167
|
-
- `/status` — AI analysis of your repo state
|
|
168
|
-
- `/diff` — Smart explanation of your changes
|
|
169
|
-
- `/commit` — Auto-generated conventional commits
|
|
170
|
-
|
|
171
|
-
</td>
|
|
172
|
-
</tr>
|
|
173
|
-
<tr>
|
|
174
|
-
<td width="50%">
|
|
175
|
-
|
|
176
|
-
### 🛡️ Enterprise Security
|
|
177
|
-
- **Permission Gate** — 4-tier command risk classification
|
|
178
|
-
- **Workspace Trust** — Per-directory security boundaries
|
|
179
|
-
- **Code Firewall** — Real-time secret/injection scanning
|
|
180
|
-
- **Audit Logging** — OWASP-aligned decision logs
|
|
181
|
-
|
|
182
|
-
</td>
|
|
183
|
-
<td width="50%">
|
|
184
|
-
|
|
185
|
-
### 🌐 Multi-Provider AI
|
|
186
|
-
- **Gemini** — Free tier, fast responses
|
|
187
|
-
- **OpenRouter** — 400+ models, one API
|
|
188
|
-
- **OpenAI** — GPT-4o, o3 reasoning
|
|
189
|
-
- **Anthropic** — Claude 4.5 Sonnet
|
|
190
|
-
- **DeepSeek** — R1, V3 models
|
|
191
|
-
- **Puter** — 100% free, no key needed
|
|
192
|
-
|
|
193
|
-
</td>
|
|
194
|
-
</tr>
|
|
195
|
-
</table>
|
|
196
|
-
|
|
197
|
-
---
|
|
198
|
-
|
|
199
|
-
## 📋 Commands Reference
|
|
200
|
-
|
|
201
|
-
### Interactive Mode (Default)
|
|
202
|
-
|
|
203
|
-
```bash
|
|
204
|
-
agdi # Start interactive coding session
|
|
205
|
-
```
|
|
206
|
-
|
|
207
|
-
Inside the session:
|
|
146
|
+
## 📋 Commands
|
|
208
147
|
|
|
209
148
|
| Command | Description |
|
|
210
149
|
|---------|-------------|
|
|
211
|
-
| `/build <prompt>` | Generate & execute an
|
|
150
|
+
| `/build <prompt>` | Generate & execute an application |
|
|
212
151
|
| `/edit <file>` | AI-powered surgical file editing |
|
|
213
|
-
| `/index` |
|
|
214
|
-
| `/search <query>` |
|
|
215
|
-
| `/
|
|
216
|
-
| `/
|
|
217
|
-
| `/memory` |
|
|
218
|
-
| `/status` | AI
|
|
219
|
-
| `/diff` | AI explanation
|
|
220
|
-
| `/commit` | Auto-generate
|
|
221
|
-
| `/clear` | Clear conversation
|
|
222
|
-
| `/history` | Show recent
|
|
152
|
+
| `/index` | Index project for semantic search |
|
|
153
|
+
| `/search <query>` | Search codebase with AI |
|
|
154
|
+
| `/agent` | Toggle multi-agent mode |
|
|
155
|
+
| `/tools` | List MCP tools |
|
|
156
|
+
| `/memory` | View persistent memory |
|
|
157
|
+
| `/status` | AI git status analysis |
|
|
158
|
+
| `/diff` | AI change explanation |
|
|
159
|
+
| `/commit` | Auto-generate commit message |
|
|
160
|
+
| `/clear` | Clear conversation |
|
|
161
|
+
| `/history` | Show recent messages |
|
|
223
162
|
| `/model` | Change AI model |
|
|
224
163
|
| `/help` | Show all commands |
|
|
225
164
|
| `/exit` | Exit Agdi |
|
|
226
165
|
|
|
227
|
-
### Direct Commands
|
|
228
|
-
|
|
229
|
-
```bash
|
|
230
|
-
agdi build "todo app" # Generate app directly
|
|
231
|
-
agdi chat # Simple chat mode
|
|
232
|
-
agdi auth # Configure API keys
|
|
233
|
-
agdi model # Change AI model
|
|
234
|
-
agdi config # Show configuration
|
|
235
|
-
agdi run [dir] # Run a generated project
|
|
236
|
-
```
|
|
237
|
-
|
|
238
166
|
---
|
|
239
167
|
|
|
240
|
-
## 🤖 Multi-Agent
|
|
241
|
-
|
|
242
|
-
Enable with `/agent` for higher quality code:
|
|
168
|
+
## 🤖 Multi-Agent Pipeline
|
|
243
169
|
|
|
244
170
|
```
|
|
245
|
-
|
|
246
|
-
│
|
|
247
|
-
|
|
248
|
-
|
|
249
|
-
|
|
250
|
-
|
|
251
|
-
│
|
|
252
|
-
│
|
|
253
|
-
|
|
254
|
-
|
|
255
|
-
|
|
256
|
-
│
|
|
257
|
-
│
|
|
258
|
-
│
|
|
259
|
-
|
|
260
|
-
|
|
261
|
-
|
|
262
|
-
│ 🔍 REVIEWER AGENT │
|
|
263
|
-
│ Reviews for: │
|
|
264
|
-
│ • Correctness • Security • Performance • Completeness │
|
|
265
|
-
│ │
|
|
266
|
-
│ → APPROVED: Code delivered to user │
|
|
267
|
-
│ → NEEDS_CHANGES: Sent back to CODER (max 3 loops) │
|
|
268
|
-
└─────────────────────────┬────────────────────────────────┘
|
|
269
|
-
▼
|
|
270
|
-
┌──────────────────────────────────────────────────────────┐
|
|
271
|
-
│ ✅ High-Quality, Reviewed Code │
|
|
272
|
-
└──────────────────────────────────────────────────────────┘
|
|
171
|
+
┌─────────────────────────────────────────────────────────────┐
|
|
172
|
+
│ 📋 PLANNER AGENT │
|
|
173
|
+
│ Analyzes request → Creates step-by-step plan │
|
|
174
|
+
└─────────────────────────────┬───────────────────────────────┘
|
|
175
|
+
▼
|
|
176
|
+
┌─────────────────────────────────────────────────────────────┐
|
|
177
|
+
│ 💻 CODER AGENT │
|
|
178
|
+
│ Writes production-ready TypeScript/React code │
|
|
179
|
+
└─────────────────────────────┬───────────────────────────────┘
|
|
180
|
+
▼
|
|
181
|
+
┌─────────────────────────────────────────────────────────────┐
|
|
182
|
+
│ 🔍 REVIEWER AGENT │
|
|
183
|
+
│ • Correctness • Security • Performance │
|
|
184
|
+
│ │
|
|
185
|
+
│ ✅ APPROVED → Delivered to user │
|
|
186
|
+
│ ⚠️ NEEDS_CHANGES → Back to CODER (max 3 loops) │
|
|
187
|
+
└─────────────────────────────────────────────────────────────┘
|
|
273
188
|
```
|
|
274
189
|
|
|
275
190
|
---
|
|
276
191
|
|
|
277
|
-
##
|
|
192
|
+
## 🛡️ Zero-Trust Security
|
|
278
193
|
|
|
279
|
-
|
|
194
|
+
### Permission Tiers
|
|
280
195
|
|
|
281
|
-
|
|
282
|
-
|
|
283
|
-
|
|
284
|
-
|
|
196
|
+
| Tier | Risk | Examples | Default |
|
|
197
|
+
|------|------|----------|---------|
|
|
198
|
+
| 0 | Read-only | `ls`, `cat`, `git status` | Auto-approve |
|
|
199
|
+
| 1 | Workspace | `touch`, `mkdir`, `git commit` | Prompt once |
|
|
200
|
+
| 2 | System | `npm install`, `docker run` | Always prompt |
|
|
201
|
+
| 3 | Dangerous | `sudo`, `rm -rf`, `curl\|bash` | Block |
|
|
285
202
|
|
|
286
|
-
|
|
287
|
-
→ /search user authentication
|
|
288
|
-
🔍 Search Results
|
|
289
|
-
src/auth/login.ts:15 (0.87)
|
|
290
|
-
export async function authenticateUser...
|
|
291
|
-
src/middleware/auth.ts:1 (0.65)
|
|
292
|
-
import { verifyToken } from...
|
|
293
|
-
src/hooks/useAuth.ts:5 (0.52)
|
|
294
|
-
const AuthContext = createContext...
|
|
295
|
-
```
|
|
203
|
+
### Security Alert (v2.6.1 Style)
|
|
296
204
|
|
|
297
|
-
**How it works:**
|
|
298
|
-
1. **Indexer** walks your project, chunks code by functions/classes
|
|
299
|
-
2. **Retriever** uses TF-IDF semantic search (no external vector DB)
|
|
300
|
-
3. **Context** is automatically injected into AI prompts
|
|
301
|
-
|
|
302
|
-
---
|
|
303
|
-
|
|
304
|
-
## 🛡️ Security Architecture
|
|
305
|
-
|
|
306
|
-
Agdi includes **zero-trust security** designed for safe AI-assisted development:
|
|
307
|
-
|
|
308
|
-
### 🚦 Permission Gate
|
|
309
|
-
|
|
310
|
-
Every command is classified by risk level:
|
|
311
|
-
|
|
312
|
-
```
|
|
313
|
-
┌─────────┬──────────────┬────────────────────────────┬──────────────┐
|
|
314
|
-
│ Tier │ Risk Level │ Examples │ Default │
|
|
315
|
-
├─────────┼──────────────┼────────────────────────────┼──────────────┤
|
|
316
|
-
│ 0 │ Read-only │ ls, cat, git status │ Auto-approve │
|
|
317
|
-
│ 1 │ Workspace │ touch, mkdir, git commit │ Prompt once │
|
|
318
|
-
│ 2 │ System │ npm install, docker run │ Always prompt│
|
|
319
|
-
│ 3 │ Dangerous │ sudo, rm -rf, curl|bash │ Block │
|
|
320
|
-
└─────────┴──────────────┴────────────────────────────┴──────────────┘
|
|
321
205
|
```
|
|
322
|
-
|
|
323
|
-
|
|
324
|
-
|
|
325
|
-
|
|
326
|
-
|
|
327
|
-
|
|
328
|
-
|
|
329
|
-
-
|
|
330
|
-
|
|
331
|
-
|
|
332
|
-
|
|
333
|
-
|
|
334
|
-
|
|
335
|
-
|
|
336
|
-
|
|
337
|
-
- Trust and remember permanently
|
|
338
|
-
- Exit without trust (read-only mode)
|
|
339
|
-
|
|
340
|
-
### 🔥 Code Firewall
|
|
341
|
-
|
|
342
|
-
Real-time scanning detects:
|
|
343
|
-
- **22+ secret patterns** — AWS, GitHub, OpenAI, Anthropic, Stripe, etc.
|
|
344
|
-
- **Shell injection** — Dangerous command patterns
|
|
345
|
-
- **Prompt injection** — LLM manipulation attempts
|
|
346
|
-
- **Network exfiltration** — Suspicious external requests
|
|
347
|
-
|
|
348
|
-
### 📊 Audit Logging
|
|
349
|
-
|
|
350
|
-
All decisions logged to `~/.agdi/audit.jsonl`:
|
|
351
|
-
|
|
352
|
-
```json
|
|
353
|
-
{"ts":"2026-01-22T00:30:00Z","action":"approve","tier":1,"cmd":"git commit","reason":"session"}
|
|
206
|
+
╔══════════════════════════════════════════════════════════════╗
|
|
207
|
+
║ 🛡️ UNTRUSTED WORKSPACE ║
|
|
208
|
+
╠══════════════════════════════════════════════════════════════╣
|
|
209
|
+
║ ║
|
|
210
|
+
║ The agent wants to execute 5 actions in this folder. ║
|
|
211
|
+
║ This includes writing files and/or running commands. ║
|
|
212
|
+
║ ║
|
|
213
|
+
║ Target: /Users/dev/projects/my-app ║
|
|
214
|
+
║ ║
|
|
215
|
+
╚══════════════════════════════════════════════════════════════╝
|
|
216
|
+
|
|
217
|
+
? Trust this workspace?
|
|
218
|
+
❯ Trust for this session
|
|
219
|
+
Trust and remember
|
|
220
|
+
Cancel
|
|
354
221
|
```
|
|
355
222
|
|
|
356
223
|
---
|
|
357
224
|
|
|
358
|
-
##
|
|
225
|
+
## 🌐 AI Providers
|
|
359
226
|
|
|
360
|
-
### Free Models (
|
|
227
|
+
### Free Models (No API Key)
|
|
361
228
|
|
|
362
229
|
| Model | ID | Best For |
|
|
363
230
|
|-------|-----|----------|
|
|
364
|
-
| GPT-OSS 120B | `openai/gpt-oss-120b:free` | General
|
|
365
|
-
|
|
|
366
|
-
|
|
|
367
|
-
| Qwen3 Next 80B | `qwen/qwen3-next-80b-a3b-instruct:free` | Complex tasks |
|
|
368
|
-
| Kimi K2 | `moonshotai/kimi-k2:free` | General purpose |
|
|
369
|
-
| Gemma 3N E2B | `google/gemma-3n-e2b-it:free` | Lightweight |
|
|
370
|
-
| LFM Thinking | `liquid/lfm-2.5-1.2b-thinking:free` | Agentic tasks |
|
|
371
|
-
|
|
372
|
-
### Image Generation
|
|
373
|
-
|
|
374
|
-
| Model | ID | Description |
|
|
375
|
-
|-------|-----|-------------|
|
|
376
|
-
| Seedream 4.5 | `bytedance-seed/seedream-4.5` | Auto-generates project images |
|
|
231
|
+
| GPT-OSS 120B | `openai/gpt-oss-120b:free` | General |
|
|
232
|
+
| Qwen3 Coder | `qwen/qwen3-coder:free` | Code |
|
|
233
|
+
| Kimi K2 | `moonshotai/kimi-k2:free` | General |
|
|
377
234
|
|
|
378
235
|
### Bring Your Own Key
|
|
379
236
|
|
|
380
|
-
|
|
381
|
-
|
|
382
|
-
|
|
383
|
-
|
|
384
|
-
|
|
385
|
-
|
|
386
|
-
- Gemini 2.5 Pro 🚀
|
|
387
|
-
- Gemini 2.0 Flash
|
|
388
|
-
|
|
389
|
-
</td>
|
|
390
|
-
<td>
|
|
391
|
-
|
|
392
|
-
**OpenRouter** 🌐
|
|
393
|
-
- Claude 4.5 Sonnet
|
|
394
|
-
- GPT-5 / GPT-4o
|
|
395
|
-
- Llama 4 Maverick
|
|
396
|
-
- DeepSeek R1
|
|
397
|
-
- 400+ more
|
|
398
|
-
|
|
399
|
-
</td>
|
|
400
|
-
<td>
|
|
401
|
-
|
|
402
|
-
**OpenAI** 🧠
|
|
403
|
-
- GPT-5
|
|
404
|
-
- GPT-4o
|
|
405
|
-
- o3
|
|
406
|
-
- o1
|
|
407
|
-
|
|
408
|
-
</td>
|
|
409
|
-
</tr>
|
|
410
|
-
<tr>
|
|
411
|
-
<td>
|
|
412
|
-
|
|
413
|
-
**Anthropic** 💜
|
|
414
|
-
- Claude 4.5 Sonnet
|
|
415
|
-
- Claude 4.5 Opus
|
|
416
|
-
- Claude 3.5 Sonnet
|
|
417
|
-
|
|
418
|
-
</td>
|
|
419
|
-
<td>
|
|
420
|
-
|
|
421
|
-
**DeepSeek** 🌊
|
|
422
|
-
- DeepSeek V3.2
|
|
423
|
-
- DeepSeek R1 (Reasoning)
|
|
424
|
-
|
|
425
|
-
</td>
|
|
426
|
-
<td>
|
|
427
|
-
|
|
428
|
-
**Get Keys** 🔑
|
|
429
|
-
- [Gemini](https://aistudio.google.com/apikey)
|
|
430
|
-
- [OpenRouter](https://openrouter.ai/keys)
|
|
431
|
-
- [OpenAI](https://platform.openai.com/api-keys)
|
|
432
|
-
- [Anthropic](https://console.anthropic.com/)
|
|
433
|
-
- [DeepSeek](https://platform.deepseek.com/)
|
|
434
|
-
|
|
435
|
-
</td>
|
|
436
|
-
</tr>
|
|
437
|
-
</table>
|
|
237
|
+
| Provider | Models | Get Key |
|
|
238
|
+
|----------|--------|---------|
|
|
239
|
+
| **Gemini** 🔥 | 2.5 Flash, 2.5 Pro | [aistudio.google.com](https://aistudio.google.com/apikey) |
|
|
240
|
+
| **OpenRouter** | Claude, GPT-4o, Llama | [openrouter.ai](https://openrouter.ai/keys) |
|
|
241
|
+
| **OpenAI** | GPT-5, o3 | [platform.openai.com](https://platform.openai.com/api-keys) |
|
|
242
|
+
| **Anthropic** | Claude 4.5 Sonnet | [console.anthropic.com](https://console.anthropic.com/) |
|
|
438
243
|
|
|
439
244
|
---
|
|
440
245
|
|
|
441
246
|
## ⚙️ Configuration
|
|
442
247
|
|
|
443
|
-
Config stored in `~/.agdi/`:
|
|
444
|
-
|
|
445
248
|
```
|
|
446
249
|
~/.agdi/
|
|
447
250
|
├── config.json # API keys & settings
|
|
448
251
|
├── rules.json # Permission rules
|
|
449
252
|
├── trusted-workspaces.json # Trusted directories
|
|
450
|
-
├── memory/ #
|
|
253
|
+
├── memory/ # Persistent memory
|
|
451
254
|
│ └── store.json
|
|
452
|
-
├── indexes/ #
|
|
255
|
+
├── indexes/ # RAG indexes
|
|
453
256
|
└── audit.jsonl # Audit log
|
|
454
257
|
```
|
|
455
258
|
|
|
456
|
-
```bash
|
|
457
|
-
# Interactive setup
|
|
458
|
-
agdi auth
|
|
459
|
-
|
|
460
|
-
# Show config
|
|
461
|
-
agdi config
|
|
462
|
-
|
|
463
|
-
# Check auth status
|
|
464
|
-
agdi auth --status
|
|
465
|
-
```
|
|
466
|
-
|
|
467
259
|
---
|
|
468
260
|
|
|
469
261
|
## 📊 Quality Metrics
|
|
470
262
|
|
|
471
|
-
<
|
|
472
|
-
<tr>
|
|
473
|
-
<td align="center">
|
|
474
|
-
<h3>0</h3>
|
|
475
|
-
<p>TypeScript Errors</p>
|
|
476
|
-
</td>
|
|
477
|
-
<td align="center">
|
|
478
|
-
<h3>60</h3>
|
|
479
|
-
<p>Tests Passing</p>
|
|
480
|
-
</td>
|
|
481
|
-
<td align="center">
|
|
482
|
-
<h3>148 KB</h3>
|
|
483
|
-
<p>Build Size</p>
|
|
484
|
-
</td>
|
|
485
|
-
<td align="center">
|
|
486
|
-
<h3>22+</h3>
|
|
487
|
-
<p>Security Patterns</p>
|
|
488
|
-
</td>
|
|
489
|
-
<td align="center">
|
|
490
|
-
<h3>900+</h3>
|
|
491
|
-
<p>Weekly Downloads</p>
|
|
492
|
-
</td>
|
|
493
|
-
</tr>
|
|
494
|
-
</table>
|
|
495
|
-
|
|
496
|
-
---
|
|
497
|
-
|
|
498
|
-
## 🚀 Examples
|
|
499
|
-
|
|
500
|
-
```bash
|
|
501
|
-
# Create a full-stack app (multi-agent recommended)
|
|
502
|
-
agdi
|
|
503
|
-
→ /agent
|
|
504
|
-
→ Create a blog with authentication and markdown support
|
|
505
|
-
|
|
506
|
-
# Semantic search your codebase
|
|
507
|
-
agdi
|
|
508
|
-
→ /index
|
|
509
|
-
→ /search payment processing
|
|
263
|
+
<div align="center">
|
|
510
264
|
|
|
511
|
-
|
|
512
|
-
|
|
513
|
-
|
|
514
|
-
|
|
265
|
+
| Metric | Value |
|
|
266
|
+
|--------|-------|
|
|
267
|
+
| TypeScript Errors | **0** |
|
|
268
|
+
| Tests Passing | **60** |
|
|
269
|
+
| Build Size | **150 KB** |
|
|
270
|
+
| Security Patterns | **22+** |
|
|
271
|
+
| Weekly Downloads | **900+** |
|
|
515
272
|
|
|
516
|
-
|
|
517
|
-
agdi
|
|
518
|
-
→ Help me debug this React component
|
|
519
|
-
→ Add dark mode to my app
|
|
520
|
-
→ Explain this TypeScript error
|
|
521
|
-
```
|
|
273
|
+
</div>
|
|
522
274
|
|
|
523
275
|
---
|
|
524
276
|
|
|
@@ -526,20 +278,18 @@ agdi
|
|
|
526
278
|
|
|
527
279
|
- **Runtime**: Node.js 18+
|
|
528
280
|
- **Language**: TypeScript 5.4
|
|
529
|
-
- **CLI
|
|
530
|
-
- **
|
|
531
|
-
- **Styling**: Chalk, Ora
|
|
281
|
+
- **CLI**: Commander, Inquirer
|
|
282
|
+
- **UI**: Chalk, Ora, Boxen, Gradient-string, Figlet
|
|
532
283
|
- **Build**: tsup
|
|
533
284
|
- **Testing**: Vitest
|
|
534
|
-
- **AI**: Multi-agent, RAG, Token Optimization
|
|
535
285
|
|
|
536
286
|
---
|
|
537
287
|
|
|
538
288
|
## 🔗 Links
|
|
539
289
|
|
|
540
|
-
- 🌐
|
|
541
|
-
- 📦
|
|
542
|
-
- 🐙
|
|
290
|
+
- 🌐 [agdi-dev.vercel.app](https://agdi-dev.vercel.app)
|
|
291
|
+
- 📦 [npmjs.com/package/agdi](https://www.npmjs.com/package/agdi)
|
|
292
|
+
- 🐙 [github.com/anassagd432/Agdi-dev](https://github.com/anassagd432/Agdi-dev)
|
|
543
293
|
|
|
544
294
|
---
|
|
545
295
|
|