@aiiware/aii 0.11.10 → 0.11.11
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 +247 -292
- package/bin/aii +432 -427
- package/package.json +4 -5
package/README.md
CHANGED
|
@@ -2,412 +2,367 @@
|
|
|
2
2
|
<img src="https://aiiware.gallerycdn.vsassets.io/extensions/aiiware/aii/0.1.1/1760456263611/Microsoft.VisualStudio.Services.Icons.Default" width="120" alt="Aii" />
|
|
3
3
|
</p>
|
|
4
4
|
|
|
5
|
-
<h1 align="center"
|
|
5
|
+
<h1 align="center">Aii — AI Agent for Your Terminal</h1>
|
|
6
6
|
|
|
7
|
-
|
|
7
|
+
<p align="center">
|
|
8
|
+
An autonomous AI assistant that lives in your terminal. Ask it anything — it reads your files, runs commands, searches the web, and writes code. No copy-pasting, no tab-switching. Just describe what you want.
|
|
9
|
+
</p>
|
|
8
10
|
|
|
9
|
-
|
|
11
|
+
<p align="center">
|
|
12
|
+
<strong>npm:</strong> <code>@aiiware/aii</code> · <strong>Version:</strong> 0.11.11 · <strong>Node.js:</strong> >= 18
|
|
13
|
+
</p>
|
|
14
|
+
|
|
15
|
+
---
|
|
16
|
+
|
|
17
|
+
## Quick Start
|
|
10
18
|
|
|
11
19
|
```bash
|
|
20
|
+
# 1. Install
|
|
12
21
|
npm install -g @aiiware/aii
|
|
22
|
+
|
|
23
|
+
# 2. Go
|
|
24
|
+
aii -D hi there!
|
|
13
25
|
```
|
|
14
26
|
|
|
15
|
-
|
|
27
|
+
On first run, Aii auto-launches a setup wizard to configure your AI provider, model, and API key. After setup completes, your command executes automatically.
|
|
16
28
|
|
|
17
|
-
|
|
18
|
-
- Aii Server running (default: `localhost:26169`)
|
|
29
|
+
Already know your provider? You can also configure directly: `aii config apikey <provider> <key>`
|
|
19
30
|
|
|
20
|
-
|
|
31
|
+
> **Requirements:** Node.js 18+ and Docker
|
|
21
32
|
|
|
22
|
-
|
|
23
|
-
# Start interactive agent mode
|
|
24
|
-
aii
|
|
33
|
+
---
|
|
25
34
|
|
|
26
|
-
|
|
27
|
-
aii "explain what this project does"
|
|
35
|
+
## What Can You Do With It?
|
|
28
36
|
|
|
29
|
-
|
|
30
|
-
cat error.log | aii "summarize the errors"
|
|
37
|
+
### Talk to it
|
|
31
38
|
|
|
32
|
-
|
|
33
|
-
aii
|
|
39
|
+
```bash
|
|
40
|
+
aii explain how this codebase is structured
|
|
34
41
|
```
|
|
35
42
|
|
|
36
|
-
|
|
37
|
-
|
|
38
|
-
Aii runs as an intelligent agent that can autonomously execute tools to complete tasks:
|
|
43
|
+
### Pipe anything in
|
|
39
44
|
|
|
40
45
|
```bash
|
|
41
|
-
aii
|
|
42
|
-
|
|
46
|
+
git diff | aii review these changes for bugs
|
|
47
|
+
cat error.log | aii what went wrong?
|
|
48
|
+
echo "Hello world" | aii translate to Japanese
|
|
43
49
|
```
|
|
44
50
|
|
|
45
|
-
###
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
|
|
49
|
-
| **Read** | Read file contents |
|
|
50
|
-
| **Write** | Create or overwrite files |
|
|
51
|
-
| **Edit** | Make targeted edits to files |
|
|
52
|
-
| **Glob** | Find files by pattern |
|
|
53
|
-
| **Grep** | Search code with regex |
|
|
54
|
-
| **Bash** | Execute shell commands |
|
|
55
|
-
| **WebSearch** | Search the web |
|
|
56
|
-
| **WebFetch** | Fetch and parse web pages |
|
|
57
|
-
| **Task** | Delegate work to specialized subagents |
|
|
58
|
-
|
|
59
|
-
### Interactive Commands
|
|
60
|
-
|
|
61
|
-
```text
|
|
62
|
-
/help Show available commands
|
|
63
|
-
/clear Clear conversation history
|
|
64
|
-
/new Start a fresh conversation
|
|
65
|
-
/history Show conversation summary
|
|
66
|
-
/model <model> Switch LLM model
|
|
67
|
-
/provider <name> Switch LLM provider
|
|
68
|
-
/context View context window usage
|
|
69
|
-
/stats Show session statistics (tokens, cost)
|
|
70
|
-
/init Generate AGENTS.md for your project
|
|
71
|
-
/memory View loaded project instructions
|
|
72
|
-
/hooks View configured hooks
|
|
73
|
-
/skills List available skills
|
|
74
|
-
/commit Generate commit message and commit
|
|
75
|
-
/loop Start autonomous agent loop
|
|
76
|
-
/mcp Manage MCP servers
|
|
77
|
-
/sessions Manage server-managed sessions
|
|
78
|
-
/agents List available subagents
|
|
79
|
-
/tasks Inspect subagent task executions
|
|
80
|
-
/diff Show recent file changes
|
|
81
|
-
/undo Undo the last file change
|
|
82
|
-
/mode Show or switch operation mode (normal/auto/plan)
|
|
83
|
-
/exit Exit the session
|
|
51
|
+
### Let it commit for you
|
|
52
|
+
|
|
53
|
+
```bash
|
|
54
|
+
aii run git commit -y
|
|
84
55
|
```
|
|
85
56
|
|
|
86
|
-
|
|
57
|
+
It reads your staged changes, writes a Conventional Commits message, and commits — one command.
|
|
87
58
|
|
|
88
|
-
|
|
89
|
-
|------------|-------------------------------------|
|
|
90
|
-
| `Enter` | Submit input |
|
|
91
|
-
| `Esc` | Cancel current request |
|
|
92
|
-
| `Ctrl+C` | Exit immediately |
|
|
93
|
-
| `Ctrl+D` | Graceful exit with session summary |
|
|
94
|
-
| `Ctrl+L` | Clear conversation |
|
|
95
|
-
| `Up/Down` | Navigate input history |
|
|
96
|
-
| `Tab` | Select from command menu |
|
|
59
|
+
### Start an interactive session
|
|
97
60
|
|
|
98
|
-
|
|
61
|
+
```bash
|
|
62
|
+
aii
|
|
63
|
+
```
|
|
99
64
|
|
|
100
|
-
|
|
65
|
+
You get a rich terminal UI with streaming responses, syntax highlighting, diff previews, and autocomplete for slash commands.
|
|
101
66
|
|
|
102
|
-
|
|
103
|
-
> analyze my codebase architecture and review the auth module
|
|
67
|
+
### Switch models on the fly
|
|
104
68
|
|
|
105
|
-
|
|
106
|
-
|
|
107
|
-
|
|
108
|
-
|
|
109
|
-
└ Running (2s)
|
|
69
|
+
```bash
|
|
70
|
+
aii --model ollama:llama3.1:8b write a bash one-liner to find large files
|
|
71
|
+
aii --model deepseek:deepseek-chat explain this regex
|
|
72
|
+
aii --model openai:gpt-4.1 refactor this function
|
|
110
73
|
```
|
|
111
74
|
|
|
112
|
-
|
|
75
|
+
Or during a session: `/model anthropic:claude-sonnet-4`
|
|
113
76
|
|
|
114
|
-
|
|
115
|
-
|-------------|----------------------------------------|
|
|
116
|
-
| **explore** | Fast codebase exploration (read-only) |
|
|
117
|
-
| **plan** | Architecture and implementation design |
|
|
118
|
-
| **review** | Code quality analysis |
|
|
77
|
+
### Run it autonomously
|
|
119
78
|
|
|
120
|
-
|
|
79
|
+
```bash
|
|
80
|
+
aii loop "fix all failing tests" --verify "npm test" --budget 0.50
|
|
81
|
+
```
|
|
121
82
|
|
|
122
|
-
|
|
83
|
+
The agent loops with fresh context each iteration, stopping when your verification command passes, the budget is hit, or you press Ctrl+C.
|
|
123
84
|
|
|
124
|
-
|
|
85
|
+
---
|
|
125
86
|
|
|
126
|
-
|
|
127
|
-
/tasks # List all task executions with IDs
|
|
128
|
-
/tasks <id> # Show full transcript for a specific task
|
|
129
|
-
```
|
|
87
|
+
## Providers
|
|
130
88
|
|
|
131
|
-
|
|
89
|
+
**8 LLM providers**, switch anytime with `aii config provider <name>`:
|
|
132
90
|
|
|
133
|
-
|
|
91
|
+
| Provider | What You Get |
|
|
92
|
+
|----------|-------------|
|
|
93
|
+
| **Anthropic** | Claude Sonnet 4, Opus 4, Haiku 4.5 |
|
|
94
|
+
| **OpenAI** | GPT-5.2, GPT-4o, GPT-4.1 |
|
|
95
|
+
| **Google** | Gemini 3-pro, 2.5-pro, 2.5-flash |
|
|
96
|
+
| **DeepSeek** | deepseek-chat, deepseek-reasoner |
|
|
97
|
+
| **Moonshot** | Kimi K2 Thinking, Vision |
|
|
98
|
+
| **Z.ai** | GLM-4.7, GLM-4.6V, GLM-4.5 |
|
|
99
|
+
| **Ollama** | Any local model — free, private, offline |
|
|
100
|
+
| **OpenRouter** | 300+ models through one API key |
|
|
134
101
|
|
|
135
|
-
|
|
136
|
-
# Pair a Telegram bot
|
|
137
|
-
aii telegram pair mybot <bot-token>
|
|
102
|
+
> **Ollama users:** `aii config provider ollama` auto-discovers your installed models and lets you pick from an interactive menu.
|
|
138
103
|
|
|
139
|
-
|
|
140
|
-
aii telegram start mybot
|
|
104
|
+
---
|
|
141
105
|
|
|
142
|
-
|
|
143
|
-
aii telegram list
|
|
144
|
-
aii telegram status mybot
|
|
145
|
-
aii telegram unpair mybot
|
|
146
|
-
```
|
|
106
|
+
## Tools
|
|
147
107
|
|
|
148
|
-
|
|
108
|
+
The agent autonomously picks from **10 built-in tools**:
|
|
149
109
|
|
|
150
|
-
|
|
110
|
+
| Tool | What It Does |
|
|
111
|
+
|------|-------------|
|
|
112
|
+
| **Read** | Read any file |
|
|
113
|
+
| **Write** | Create or overwrite files |
|
|
114
|
+
| **Edit** | Surgical find-and-replace edits |
|
|
115
|
+
| **Glob** | Find files by pattern |
|
|
116
|
+
| **Grep** | Search file contents (regex) |
|
|
117
|
+
| **Bash** | Run shell commands |
|
|
118
|
+
| **WebSearch** | Search the web |
|
|
119
|
+
| **WebFetch** | Fetch and parse web pages |
|
|
120
|
+
| **Task** | Delegate to parallel subagents |
|
|
121
|
+
| **Skill** | Invoke extensible skill packages |
|
|
151
122
|
|
|
152
|
-
|
|
123
|
+
You control what the agent can do. By default, it asks before running commands or writing files. Use `--auto` to skip confirmations.
|
|
153
124
|
|
|
154
|
-
|
|
155
|
-
# List recent sessions
|
|
156
|
-
aii sessions list
|
|
125
|
+
---
|
|
157
126
|
|
|
158
|
-
|
|
159
|
-
|
|
127
|
+
## Slash Commands
|
|
128
|
+
|
|
129
|
+
During interactive sessions, type `/` to see all commands. Highlights:
|
|
130
|
+
|
|
131
|
+
| Command | What It Does |
|
|
132
|
+
|---------|-------------|
|
|
133
|
+
| `/help` | Show all commands |
|
|
134
|
+
| `/model [name]` | View or switch model |
|
|
135
|
+
| `/provider [name]` | View or switch provider |
|
|
136
|
+
| `/auto` | Toggle auto-approve mode |
|
|
137
|
+
| `/plan` | Toggle plan mode (read-only) |
|
|
138
|
+
| `/diff` | See all file changes this session |
|
|
139
|
+
| `/undo` | Revert the last file change |
|
|
140
|
+
| `/stats` | Session stats — tokens, cost, duration |
|
|
141
|
+
| `/context` | Context window usage; `/context clear` to summarize |
|
|
142
|
+
| `/commit` | AI-powered git commit |
|
|
143
|
+
| `/review-pr` | AI-powered PR review |
|
|
144
|
+
| `/worktree create` | Isolated workspace for experiments |
|
|
145
|
+
| `/mcp install` | Add MCP server from catalog |
|
|
146
|
+
| `/skills` | List and manage skill packages |
|
|
147
|
+
| `/sessions` | Browse and resume past sessions |
|
|
148
|
+
| `/exit` | End session with summary |
|
|
160
149
|
|
|
161
|
-
|
|
162
|
-
aii sessions info <id>
|
|
150
|
+
### Keyboard Shortcuts
|
|
163
151
|
|
|
164
|
-
|
|
165
|
-
|
|
166
|
-
|
|
167
|
-
|
|
152
|
+
| Key | Action |
|
|
153
|
+
|-----|--------|
|
|
154
|
+
| `Enter` | Submit input |
|
|
155
|
+
| `Esc` | Cancel current request |
|
|
156
|
+
| `Ctrl+C` | Exit immediately |
|
|
157
|
+
| `Ctrl+D` | Graceful exit with session summary |
|
|
158
|
+
| `Ctrl+L` | Clear conversation |
|
|
159
|
+
| `Up/Down` | Navigate input history |
|
|
160
|
+
| `Tab` | Autocomplete commands |
|
|
168
161
|
|
|
169
|
-
|
|
162
|
+
---
|
|
170
163
|
|
|
171
|
-
|
|
164
|
+
## Key Features
|
|
172
165
|
|
|
173
|
-
|
|
174
|
-
# In Telegram: work on a task via bot
|
|
175
|
-
# Later, from CLI:
|
|
176
|
-
aii sessions list --channel telegram
|
|
177
|
-
aii sessions resume abc123
|
|
178
|
-
```
|
|
166
|
+
### Subagents
|
|
179
167
|
|
|
180
|
-
|
|
168
|
+
The agent delegates complex tasks to specialized subagents running in parallel:
|
|
181
169
|
|
|
182
|
-
|
|
183
|
-
|
|
184
|
-
|
|
185
|
-
|
|
186
|
-
|
|
187
|
-
|
|
188
|
-
```
|
|
170
|
+
| Agent | Purpose | Tools |
|
|
171
|
+
|-------|---------|-------|
|
|
172
|
+
| **Explore** | Fast codebase search and analysis | Read-only |
|
|
173
|
+
| **Plan** | Architecture and implementation design | Read-only |
|
|
174
|
+
| **Review** | Code quality analysis | Read-only |
|
|
175
|
+
| **Code** | Full implementation | All tools |
|
|
189
176
|
|
|
190
|
-
|
|
177
|
+
Define custom agents in `.aii/agents/<name>.md` with YAML frontmatter for custom system prompts, model tiers, and tool access.
|
|
191
178
|
|
|
192
|
-
|
|
179
|
+
### Git Worktrees
|
|
193
180
|
|
|
194
|
-
|
|
181
|
+
Work on multiple features in parallel without stashing:
|
|
195
182
|
|
|
196
183
|
```bash
|
|
197
|
-
|
|
198
|
-
|
|
199
|
-
|
|
200
|
-
|
|
201
|
-
/mcp remove myserver # Remove a server
|
|
184
|
+
aii worktree create auth-refactor # Create isolated workspace + branch
|
|
185
|
+
aii worktree list # See all worktrees
|
|
186
|
+
aii worktree merge auth-refactor # Merge back
|
|
187
|
+
aii worktree clean # Remove all clean worktrees
|
|
202
188
|
```
|
|
203
189
|
|
|
204
|
-
|
|
190
|
+
Or use `/worktree` slash commands during a session.
|
|
205
191
|
|
|
206
|
-
|
|
192
|
+
### Autonomous Loops
|
|
207
193
|
|
|
208
|
-
Run the agent in a continuous loop
|
|
194
|
+
Run the agent in a continuous loop with guardrails:
|
|
209
195
|
|
|
210
196
|
```bash
|
|
211
|
-
|
|
212
|
-
aii loop --verify "npm test"
|
|
213
|
-
|
|
214
|
-
# Or start within a session
|
|
215
|
-
/loop --max 10 --promise "All tests passing"
|
|
197
|
+
aii loop "fix all TypeScript errors"
|
|
198
|
+
aii loop --max 10 --verify "npm test" "add missing tests"
|
|
199
|
+
aii loop --budget 1.00 --duration 600 "refactor the auth module"
|
|
216
200
|
```
|
|
217
201
|
|
|
218
|
-
**
|
|
202
|
+
**Exit conditions:** `--max N`, `--budget $N`, `--verify "cmd"`, `--promise "text"`, `--duration <sec>`, `--input-tokens N`, `--output-tokens N`
|
|
219
203
|
|
|
220
|
-
|
|
221
|
-
|----------------------|------------------------------------------|
|
|
222
|
-
| `--max <N>` | Maximum iterations |
|
|
223
|
-
| `--verify "<cmd>"` | Run verification command after each step |
|
|
224
|
-
| `--promise "<text>"` | Stop when output contains this text |
|
|
225
|
-
| `--budget <$N>` | Cost limit |
|
|
226
|
-
| `--duration <time>` | Time limit (e.g., `30m`, `2h`) |
|
|
204
|
+
### Telegram Integration
|
|
227
205
|
|
|
228
|
-
|
|
229
|
-
|
|
230
|
-
Aii supports multiple LLM providers. The provider is auto-detected from the model name:
|
|
231
|
-
|
|
232
|
-
| Provider | Models | Setup |
|
|
233
|
-
| --- | --- | --- |
|
|
234
|
-
| **Anthropic** | Claude Sonnet, Opus, Haiku | API key |
|
|
235
|
-
| **OpenAI** | GPT-4o, GPT-4, o1, o3 | API key |
|
|
236
|
-
| **Google** | Gemini Pro, Flash | API key |
|
|
237
|
-
| **DeepSeek** | DeepSeek Chat, Coder | API key |
|
|
238
|
-
| **Moonshot** | Kimi K2 | API key |
|
|
239
|
-
| **Zai** | GLM-4 | API key |
|
|
240
|
-
| **Ollama** | Llama, Mistral, and more | Local, no API key |
|
|
241
|
-
| **OpenRouter** | 300+ models via single API key | API key |
|
|
206
|
+
Run the full agent from Telegram:
|
|
242
207
|
|
|
243
208
|
```bash
|
|
244
|
-
#
|
|
245
|
-
|
|
246
|
-
|
|
247
|
-
|
|
248
|
-
# Use provider:model format for one-shot commands
|
|
249
|
-
aii "explain this code" --model deepseek:deepseek-chat
|
|
250
|
-
aii "translate hello" --model ollama:gpt-oss:20b
|
|
209
|
+
aii telegram pair mybot <bot-token> # Pair with @BotFather token
|
|
210
|
+
aii telegram start mybot # Start listening
|
|
211
|
+
aii telegram list # Manage bots
|
|
251
212
|
```
|
|
252
213
|
|
|
253
|
-
|
|
254
|
-
|
|
255
|
-
## Context Window Management (v0.3)
|
|
214
|
+
Start a conversation in Telegram, continue it from the CLI — or vice versa.
|
|
256
215
|
|
|
257
|
-
|
|
216
|
+
### Skills System
|
|
258
217
|
|
|
259
|
-
|
|
260
|
-
- **Auto-summarization** triggers when approaching the limit, preserving recent context
|
|
261
|
-
- **Dynamic thresholds** adjust per model's output capacity
|
|
218
|
+
Extensible instruction packages you can install and invoke:
|
|
262
219
|
|
|
263
220
|
```bash
|
|
264
|
-
/
|
|
265
|
-
/
|
|
221
|
+
/skills # List all skills
|
|
222
|
+
/skills install owner/repo/skill-name # Install from GitHub
|
|
223
|
+
/commit # Built-in: AI git commit
|
|
224
|
+
/review-pr # Built-in: PR review
|
|
225
|
+
/<custom-skill> [args] # Any user-invocable skill
|
|
266
226
|
```
|
|
267
227
|
|
|
268
|
-
|
|
228
|
+
Create skills in `.aii/skills/<name>/SKILL.md` with YAML frontmatter.
|
|
229
|
+
|
|
230
|
+
### Project Instructions (AGENTS.md)
|
|
269
231
|
|
|
270
|
-
|
|
232
|
+
Drop an `AGENTS.md` file in your project root to give the agent project-specific context — coding conventions, architecture notes, preferred patterns. It's loaded automatically.
|
|
271
233
|
|
|
272
234
|
```bash
|
|
273
235
|
/init # Auto-generate by analyzing your project
|
|
274
|
-
/init --with-hooks # Also generate hooks.json template
|
|
275
236
|
/memory # View loaded instructions
|
|
276
|
-
/memory reload # Hot-reload without restart
|
|
277
237
|
```
|
|
278
238
|
|
|
279
|
-
|
|
239
|
+
Also discovers `CLAUDE.md`, `.cursorrules`, and `.github/copilot-instructions.md`.
|
|
280
240
|
|
|
281
|
-
|
|
241
|
+
### Hook System
|
|
282
242
|
|
|
283
|
-
Run custom scripts
|
|
243
|
+
Run custom scripts when the agent takes actions. Configure in `.aii/hooks.json`:
|
|
244
|
+
|
|
245
|
+
**Events:** `PreToolUse`, `PostToolUse`, `Stop`, `SubagentStart`, `SubagentStop`
|
|
284
246
|
|
|
285
247
|
```bash
|
|
286
|
-
/init --with-hooks # Generate hooks template
|
|
287
248
|
/hooks # View configured hooks
|
|
288
249
|
/hooks reload # Hot-reload without restart
|
|
289
250
|
```
|
|
290
251
|
|
|
291
|
-
|
|
292
|
-
|
|
293
|
-
## Skills (v0.6)
|
|
294
|
-
|
|
295
|
-
Extend the agent with reusable skill packages:
|
|
296
|
-
|
|
297
|
-
**Natural language** — the agent auto-detects the right skill:
|
|
298
|
-
|
|
299
|
-
```text
|
|
300
|
-
> commit my changes
|
|
301
|
-
> review the latest PR
|
|
302
|
-
```
|
|
252
|
+
### MCP Servers
|
|
303
253
|
|
|
304
|
-
|
|
254
|
+
Extend the agent with Model Context Protocol servers:
|
|
305
255
|
|
|
306
256
|
```bash
|
|
307
|
-
|
|
308
|
-
|
|
257
|
+
aii mcp catalog # Browse available servers
|
|
258
|
+
aii mcp install github # One-command install
|
|
259
|
+
aii mcp add my-server npx my-mcp # Add custom server
|
|
260
|
+
aii mcp list-tools github # Discover tools
|
|
261
|
+
aii mcp test github # Connectivity test
|
|
309
262
|
```
|
|
310
263
|
|
|
311
|
-
|
|
264
|
+
### Server Sessions
|
|
265
|
+
|
|
266
|
+
Conversations persist across restarts and channels:
|
|
312
267
|
|
|
313
268
|
```bash
|
|
314
|
-
|
|
315
|
-
|
|
316
|
-
|
|
317
|
-
/skills uninstall code-review # Remove a skill
|
|
318
|
-
/skills reload # Re-discover skills
|
|
269
|
+
aii sessions list # Browse recent sessions
|
|
270
|
+
aii sessions resume <id> # Pick up where you left off
|
|
271
|
+
aii sessions info <id> # View stats (tokens, cost, tools)
|
|
319
272
|
```
|
|
320
273
|
|
|
321
|
-
**Creating skills** — add `SKILL.md` files in `.aii/skills/`:
|
|
322
|
-
|
|
323
|
-
```markdown
|
|
324
|
-
---
|
|
325
|
-
name: my-skill
|
|
326
|
-
description: Does something useful
|
|
327
|
-
allowed-tools: Bash Read Grep
|
|
328
274
|
---
|
|
329
275
|
|
|
330
|
-
Instructions for the agent...
|
|
331
|
-
Use $ARGUMENTS for user input.
|
|
332
|
-
```
|
|
333
|
-
|
|
334
276
|
## Configuration
|
|
335
277
|
|
|
336
278
|
```bash
|
|
337
|
-
aii config
|
|
338
|
-
aii config
|
|
339
|
-
aii config
|
|
340
|
-
aii config
|
|
341
|
-
aii config provider
|
|
279
|
+
aii config init # Interactive setup wizard
|
|
280
|
+
aii config show # Current settings
|
|
281
|
+
aii config validate # Check config + API key validity
|
|
282
|
+
aii config set <key> <value> # Set a specific value
|
|
283
|
+
aii config provider [name] # Show or switch provider
|
|
284
|
+
aii config model [name] # Show or switch model
|
|
285
|
+
aii config apikey <provider> [key] # Set API key (masked input)
|
|
286
|
+
aii config apikey list # Show all provider key status
|
|
287
|
+
aii config apikey remove <provider> # Remove a provider's key
|
|
288
|
+
aii config apikey help # Show API key help
|
|
289
|
+
aii config host [host:port] # Show or change server URL
|
|
290
|
+
aii config models # List all available models
|
|
291
|
+
aii config providers # List all providers with status
|
|
292
|
+
aii config help # Show config help
|
|
342
293
|
```
|
|
343
294
|
|
|
344
|
-
|
|
295
|
+
Config lives in `~/.aii/`:
|
|
345
296
|
|
|
346
|
-
|
|
297
|
+
| File | Purpose |
|
|
298
|
+
|------|---------|
|
|
299
|
+
| `config.yaml` | Provider, model, server host, preferences |
|
|
300
|
+
| `secrets.yaml` | API keys (file mode 0600) |
|
|
301
|
+
| `mcp_servers.json` | MCP server configuration |
|
|
347
302
|
|
|
348
|
-
|
|
349
|
-
llm:
|
|
350
|
-
provider: anthropic
|
|
351
|
-
model: claude-sonnet-4-6
|
|
352
|
-
temperature: 0.7
|
|
303
|
+
Environment variables also work: `ANTHROPIC_API_KEY`, `OPENAI_API_KEY`, `DEEPSEEK_API_KEY`, etc.
|
|
353
304
|
|
|
354
|
-
|
|
355
|
-
url: http://localhost:26169
|
|
356
|
-
```
|
|
305
|
+
---
|
|
357
306
|
|
|
358
|
-
|
|
307
|
+
## CLI Reference
|
|
308
|
+
|
|
309
|
+
| Command | Description |
|
|
310
|
+
|---------|-------------|
|
|
311
|
+
| `aii` | Start interactive agent session |
|
|
312
|
+
| `aii <query>` | Single-shot query |
|
|
313
|
+
| `aii run git commit [-y]` | AI-powered git commit |
|
|
314
|
+
| `aii loop [query] [--flags]` | Autonomous agent loop |
|
|
315
|
+
| `aii config <action>` | Configuration management |
|
|
316
|
+
| `aii mcp <action>` | MCP server management |
|
|
317
|
+
| `aii telegram <action>` | Telegram bot management |
|
|
318
|
+
| `aii sessions <action>` | Session management |
|
|
319
|
+
| `aii worktree <action>` | Git worktree management |
|
|
320
|
+
| `aii prompt <action>` | Prompt library |
|
|
321
|
+
| `aii doctor` | Health diagnostics |
|
|
322
|
+
| `aii help` | Show help |
|
|
323
|
+
| `aii version` | Show version |
|
|
324
|
+
|
|
325
|
+
### Global Flags
|
|
326
|
+
|
|
327
|
+
| Flag | Description |
|
|
328
|
+
|------|-------------|
|
|
329
|
+
| `--model <name>` | Override model (e.g., `deepseek:deepseek-chat`) |
|
|
330
|
+
| `--auto` | Auto-approve all tool actions |
|
|
331
|
+
| `-y`, `--yes` | Auto-confirm prompts |
|
|
332
|
+
| `-c`, `--continue-chat <id>` | Resume a specific session |
|
|
333
|
+
| `-n`, `--new-chat` | Force new session |
|
|
334
|
+
| `-i`, `--interactive` | Force interactive mode |
|
|
335
|
+
| `--offline` | Disable web search and MCP |
|
|
336
|
+
| `-D`, `--direct` | Direct LLM mode (bypass server intent recognition) |
|
|
337
|
+
| `--no-colors` | Disable colors |
|
|
338
|
+
| `--no-streaming` | Disable streaming output |
|
|
339
|
+
| `--show-tokens` | Show token usage |
|
|
340
|
+
| `--show-cost` | Show cost estimates |
|
|
341
|
+
| `-v`, `--verbose` | Verbose output |
|
|
342
|
+
| `-d`, `--debug` | Debug-level output |
|
|
359
343
|
|
|
360
|
-
|
|
361
|
-
anthropic_api_key: sk-ant-...
|
|
362
|
-
openai_api_key: sk-...
|
|
363
|
-
deepseek_api_key: sk-...
|
|
364
|
-
```
|
|
344
|
+
---
|
|
365
345
|
|
|
366
|
-
|
|
346
|
+
## Troubleshooting
|
|
367
347
|
|
|
368
|
-
|
|
348
|
+
Run `aii doctor` for automated diagnostics.
|
|
369
349
|
|
|
370
|
-
|
|
|
371
|
-
|
|
372
|
-
|
|
|
373
|
-
|
|
|
374
|
-
|
|
|
375
|
-
|
|
|
376
|
-
|
|
|
377
|
-
|
|
|
378
|
-
| `--clean` | Clean output mode |
|
|
379
|
-
| `--standard` | Standard output mode (default) |
|
|
380
|
-
| `--thinking` | Show full reasoning |
|
|
381
|
-
| `--no-streaming` | Disable streaming output |
|
|
382
|
-
| `-q, --quiet` | Minimal output |
|
|
383
|
-
| `-y, --yes` | Auto-confirm shell commands |
|
|
350
|
+
| Problem | Fix |
|
|
351
|
+
|---------|-----|
|
|
352
|
+
| `aii` not found | `npm install -g @aiiware/aii`, restart terminal |
|
|
353
|
+
| Server won't start | Make sure Docker is running |
|
|
354
|
+
| API key error | `aii config apikey <provider> <key>` or `aii config init` |
|
|
355
|
+
| Wrong model | `aii config model <name>` |
|
|
356
|
+
| Ollama not connecting | Start Ollama first, then `aii config provider ollama` |
|
|
357
|
+
| Context window full | `/context clear` during session |
|
|
384
358
|
|
|
385
|
-
|
|
359
|
+
---
|
|
386
360
|
|
|
387
|
-
|
|
388
|
-
|
|
389
|
-
|
|
390
|
-
|
|
391
|
-
|
|
392
|
-
|
|
393
|
-
```
|
|
361
|
+
## Links
|
|
362
|
+
|
|
363
|
+
- **npm**: [@aiiware/aii](https://www.npmjs.com/package/@aiiware/aii)
|
|
364
|
+
- **Website**: [aiiware.com](https://aiiware.com)
|
|
365
|
+
|
|
366
|
+
---
|
|
394
367
|
|
|
395
|
-
|
|
396
|
-
|
|
397
|
-
- **Autonomous Agent** — Executes multi-step tasks with tool use
|
|
398
|
-
- **Multi-Agent Delegation** — Parallel subagents for complex tasks
|
|
399
|
-
- **Telegram Access** — Use your agent remotely via Telegram bots
|
|
400
|
-
- **MCP Integration** — Extend with external tools via Model Context Protocol
|
|
401
|
-
- **Autonomous Loops** — Continuous agent execution with verification and budgets
|
|
402
|
-
- **Multiple Providers** — Claude, GPT, Gemini, DeepSeek, Ollama, OpenRouter
|
|
403
|
-
- **Real-time Streaming** — Token-by-token response display
|
|
404
|
-
- **Context Management** — Auto-summarization prevents token limit errors
|
|
405
|
-
- **Project Instructions** — Teach the agent your conventions with AGENTS.md
|
|
406
|
-
- **Skills System** — Extend with reusable skills, invokable via commands or natural language
|
|
407
|
-
- **Hooks** — Custom scripts for validation, logging, and approval gates
|
|
408
|
-
- **Server Sessions** — Persistent conversations that survive restarts and resume across channels
|
|
409
|
-
- **Session Persistence** — Continue previous conversations with `--continue-chat` or `aii sessions resume`
|
|
410
|
-
|
|
411
|
-
## License
|
|
412
|
-
|
|
413
|
-
Proprietary - Copyright 2025-present AiiWare.com. All Rights Reserved.
|
|
368
|
+
**License:** Proprietary — Copyright 2025-present AiiWare.com. All rights reserved.
|