@astramindapp/mcp-server 0.5.0 → 0.5.1

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.
Files changed (2) hide show
  1. package/README.md +212 -139
  2. package/package.json +1 -1
package/README.md CHANGED
@@ -1,19 +1,28 @@
1
1
  # @astramindapp/mcp-server
2
2
 
3
- **MIND MCP Server** — Give any AI agent persistent memory.
4
-
5
- Your AI agents forget everything between sessions. MIND fixes that. Connect any MCP-compatible tool to your personal knowledge graph and never lose context again.
6
-
7
- ## What It Does
8
-
9
- | Tool | Description |
10
- |------|-------------|
11
- | `mind_query` | Search your knowledge graph — returns only what's relevant |
12
- | `mind_remember` | Store anything — auto-categorized as document, entry, or thought |
13
- | `mind_context` | Load your persistent identity, preferences, rules, and priorities |
14
- | `mind_life` | Manage goals, projects, and tasks |
15
- | `mind_crm` | Track contacts and relationships |
16
- | `mind_graph` | Check graph health and statistics |
3
+ **MIND MCP Server** — The most complete AI memory layer available. 15 tools, 89 actions.
4
+
5
+ Your AI agents forget everything between sessions. MIND fixes that. Connect any MCP-compatible agent to your personal knowledge graph — with emotional intelligence, CRM, life management, social features, self-training, autonomous insights, and more.
6
+
7
+ ## 15 Tools
8
+
9
+ | Tool | Actions | What It Does |
10
+ |------|---------|-------------|
11
+ | `mind_query` | 1 | Semantic search across your knowledge graph (5 search modes) |
12
+ | `mind_remember` | 5 | Store, search, get, list, delete — documents, entries, thoughts |
13
+ | `mind_context` | 1 | Load persistent identity, preferences, rules, priorities, recent activity |
14
+ | `mind_life` | 12 | Goals, projects, tasks + full calendar management + productivity stats |
15
+ | `mind_crm` | 7 | Contacts, pipeline, activity logging, interaction history |
16
+ | `mind_graph` | 3 | Graph stats, diagnostics, entity labels |
17
+ | `mind_admin` | 7 | User provisioning, featured minds, tier/credit management |
18
+ | `mind_sense` | 7 | MINDsense emotional intelligence — state, signals, timeline, KG weights, spikes |
19
+ | `mind_research` | 3 | Launch autonomous deep research jobs |
20
+ | `mind_train` | 7 | Self-training sessions + save chats to knowledge graph |
21
+ | `mind_social` | 14 | Thoughts (posts), social feed, communities, likes, comments |
22
+ | `mind_profile` | 9 | Profile, custom system prompts, LLM model selection |
23
+ | `mind_insights` | 7 | Autonomous Learning Engine insights, weekly summaries, feedback |
24
+ | `mind_automate` | 6 | Scheduled automations, event triggers, execution history |
25
+ | `mind_notify` | 4 | Notifications, mark read, stats |
17
26
 
18
27
  ## Quick Start
19
28
 
@@ -29,6 +38,12 @@ npm install -g @astramindapp/mcp-server
29
38
 
30
39
  ### 3. Configure Your AI Tool
31
40
 
41
+ #### Claude Code
42
+
43
+ ```bash
44
+ claude mcp add mind -- env MIND_API_KEY=mind_xxx mind-mcp
45
+ ```
46
+
32
47
  #### Claude Desktop
33
48
 
34
49
  Add to your `claude_desktop_config.json`:
@@ -46,29 +61,6 @@ Add to your `claude_desktop_config.json`:
46
61
  }
47
62
  ```
48
63
 
49
- #### Claude Code
50
-
51
- ```bash
52
- claude mcp add mind -- env MIND_API_KEY=mind_xxx mind-mcp
53
- ```
54
-
55
- #### OpenClaw
56
-
57
- Add to your OpenClaw config:
58
-
59
- ```json5
60
- {
61
- mcp: {
62
- servers: {
63
- mind: {
64
- command: "mind-mcp",
65
- env: { MIND_API_KEY: "mind_xxx" }
66
- }
67
- }
68
- }
69
- }
70
- ```
71
-
72
64
  #### Cursor
73
65
 
74
66
  Add to `.cursor/mcp.json`:
@@ -86,133 +78,203 @@ Add to `.cursor/mcp.json`:
86
78
  }
87
79
  ```
88
80
 
89
- ### `mind_admin` — Partner User Provisioning
81
+ #### Any MCP-Compatible Agent
90
82
 
91
- Admin-only operations. Requires an admin API key.
83
+ The server works with any tool that supports Model Context Protocol — Claude, GPT, Gemini, Llama, or any future model.
92
84
 
93
- | Action | Description |
94
- |--------|-------------|
95
- | `create_user` | Provision a new MIND account (for partner app integration) |
96
- | `list_users` | List all users with analytics |
97
- | `update_user_tier` | Change a user's subscription tier |
98
- | `adjust_user_credits` | Add or deduct credits |
99
- | `create_featured_mind` | Create a public featured mind |
100
- | `list_featured_minds` | List featured minds catalog |
101
- | `update_featured_mind` | Update a featured mind |
102
-
103
- **Partner User Provisioning** (`create_user`): Programmatically create MIND accounts when users sign up for your app. Returns a JWT and optionally a long-lived developer API key.
85
+ ## How It Works
104
86
 
105
87
  ```
106
- username: "newuser123"
107
- email: "user@partnerapp.com"
108
- password: "securepassword"
109
- source: "myapp" # optional - tracks which partner app created the user
110
- tier: "free" # optional - free, starter, professional, business
111
- generate_api_key: true # optional - generates a permanent API key (mind_...) for this user
112
- api_key_name: "myapp" # optional - display name for the API key
88
+ Your AI Agent <--> MCP Protocol <--> MIND MCP Server <--> Personal Knowledge Graph
89
+ |
90
+ Emotional Intelligence
91
+ Autonomous Learning
92
+ CRM + Life + Social
113
93
  ```
114
94
 
115
- > Set `generate_api_key: true` to get a permanent API key in the response — your app stores it and uses it for all subsequent MIND API calls on behalf of this user.
95
+ 1. Agent calls `mind_context` at session start → loads identity, rules, priorities
96
+ 2. Agent calls `mind_query` before decisions → retrieves relevant memories
97
+ 3. Agent calls `mind_remember` after tasks → stores outcomes and learnings
98
+ 4. Agent calls `mind_sense` → reads user's emotional state to adapt responses
99
+ 5. Agent calls `mind_insights` → surfaces autonomous pattern detection
100
+ 6. Next session, any agent has full context. Knowledge compounds.
116
101
 
117
- ## Partner Keys
102
+ ## What Makes MIND Different
118
103
 
119
- Partner keys are scoped credentials that can **only** provision new users — they cannot access any other admin endpoint. Create them from your admin account:
104
+ ### vs. Flat File Memory (MEMORY.md)
105
+
106
+ | | Flat Files | MIND |
107
+ |---|---|---|
108
+ | Size | ~20K chars, truncated | Unlimited knowledge graph |
109
+ | Retrieval | Loads everything every turn | Only relevant memories via semantic search |
110
+ | Structure | Unstructured text | Graph with entities, relationships, and emotional weights |
111
+ | Cross-agent | One tool only | Shared across all AI agents |
112
+ | Intelligence | None | Autonomous pattern detection + emotional encoding |
113
+
114
+ ### vs. Other MCP Memory Servers
115
+
116
+ | | Anthropic Official | Mem0 | Graphiti/Zep | MIND |
117
+ |---|---|---|---|---|
118
+ | Tools | 8 | 4 | 9 | **15** |
119
+ | Knowledge graph | Basic (JSON) | No (vectors) | Yes (Neo4j) | **Yes (LightRAG)** |
120
+ | Emotional intelligence | No | No | No | **Yes (patented)** |
121
+ | CRM | No | No | No | **Yes** |
122
+ | Life management | No | No | No | **Yes** |
123
+ | Social features | No | No | No | **Yes** |
124
+ | Self-training | No | No | No | **Yes** |
125
+ | Research agent | No | No | No | **Yes** |
126
+ | Automations | No | No | No | **Yes** |
127
+ | Mobile app | No | No | No | **Yes** |
128
+
129
+ ## Tool Reference
130
+
131
+ ### `mind_query` — Search Your Knowledge Graph
120
132
 
121
- ```bash
122
- curl -X POST https://m-i-n-d.ai/admin/partner-keys \
123
- -H "X-API-Key: mind_YOUR_ADMIN_KEY" \
124
- -H "Content-Type: application/json" \
125
- -d '{"partner_name": "mypartnerapp"}'
133
+ ```
134
+ query: "What did I decide about the authentication approach?"
135
+ mode: "hybrid" // hybrid (default), mix, global, local, naive
126
136
  ```
127
137
 
128
- The response includes a `mind_partner_...` key. Give this to your partner developer — it's the only credential they need.
138
+ Returns an AI-synthesized answer from your stored documents, entries, and thoughts with source attribution.
129
139
 
130
- | Endpoint | Method | Description |
131
- |----------|--------|-------------|
132
- | `/admin/partner-keys` | POST | Create a partner key (admin-only) |
133
- | `/admin/partner-keys` | GET | List all partner keys (admin-only) |
134
- | `/admin/partner-keys/{id}` | DELETE | Revoke a partner key (admin-only) |
140
+ ### `mind_remember` — Store & Manage Content
135
141
 
136
- ## Admin REST API: User Provisioning
142
+ | Action | Description |
143
+ |--------|-------------|
144
+ | `create` | Store content (auto-categorized as document, entry, or thought) |
145
+ | `delete` | Remove by ID |
146
+ | `search` | Find entries/thoughts by query |
147
+ | `get` | Retrieve specific item by ID |
148
+ | `list` | Paginated listing of all content |
137
149
 
138
- Partner apps create MIND accounts using their partner key (or admin key) via REST:
150
+ ### `mind_context` — Load Persistent Context
139
151
 
140
- ```bash
141
- curl -X POST https://m-i-n-d.ai/admin/users/create \
142
- -H "X-API-Key: mind_partner_YOUR_PARTNER_KEY" \
143
- -H "Content-Type: application/json" \
144
- -d '{
145
- "username": "newuser123",
146
- "email": "user@partnerapp.com",
147
- "password": "securepassword",
148
- "tier": "free",
149
- "generate_api_key": true,
150
- "api_key_name": "myapp integration"
151
- }'
152
- ```
152
+ Loads five structured sections at session start:
153
+ - **Soul** — Core identity, mission, personality
154
+ - **User** — Who the user is, their role, preferences
155
+ - **Rules** — Operating constraints, behavioral guidelines
156
+ - **Priorities** — Current goals, active projects, deadlines
157
+ - **Recent** — Latest activity, outcomes, decisions
153
158
 
154
- > When using a partner key, the `source` is automatically attached from the key's configuration. Set `generate_api_key: true` to get a permanent developer API key in the same response — no second call needed.
159
+ ### `mind_life` — Life Management + Calendar
155
160
 
156
- **Response:**
157
- ```json
158
- {
159
- "status": "success",
160
- "access_token": "eyJ...",
161
- "token_type": "bearer",
162
- "user": {
163
- "username": "newuser123",
164
- "email": "user@partnerapp.com",
165
- "workspace_id": "user_newuser123_a1b2c3d4",
166
- "tier": "free",
167
- "source": "myapp"
168
- },
169
- "api_key": {
170
- "id": "uuid",
171
- "name": "myapp integration",
172
- "key": "mind_aBcDeFgH...",
173
- "prefix": "mind_aBcDeFg",
174
- "scopes": ["query", "documents:read", "documents:write", "...all 19 scopes"]
175
- }
176
- }
177
- ```
161
+ | Action | Description |
162
+ |--------|-------------|
163
+ | `list`, `create`, `update`, `complete`, `delete`, `move`, `get` | Full task/goal CRUD |
164
+ | `calendar_list`, `calendar_create`, `calendar_update`, `calendar_delete` | Calendar events |
165
+ | `stats` | Productivity metrics and completion rates |
178
166
 
179
- | Field | Type | Required | Description |
180
- |-------|------|----------|-------------|
181
- | `username` | string | Yes | 3-30 chars, letters/numbers/underscores |
182
- | `email` | string | Yes | Valid email address |
183
- | `password` | string | Yes | 8-128 characters |
184
- | `source` | string | No | Your app/partner identifier |
185
- | `tier` | string | No | Subscription tier (default: `free`) |
186
- | `generate_api_key` | boolean | No | Generate a permanent API key for this user (default: `false`) |
187
- | `api_key_name` | string | No | Display name for the API key |
167
+ ### `mind_crm` — Contact Relationship Management
188
168
 
189
- The `api_key.key` is a permanent developer key (no expiry) — store it and use it for all `/developer/v1/...` calls on behalf of this user. The `access_token` (JWT) expires in 48 hours.
169
+ | Action | Description |
170
+ |--------|-------------|
171
+ | `list`, `create`, `update`, `delete`, `get` | Contact CRUD with pipeline stages |
172
+ | `log_activity` | Record calls, emails, meetings, notes |
173
+ | `list_activities` | View interaction history |
190
174
 
191
- This endpoint accepts either a **partner key** or an **admin API key**. Unlike the public `/register` endpoint, it bypasses email uniqueness restrictions so partner apps can freely provision accounts.
175
+ ### `mind_sense` — MINDsense Emotional Intelligence
192
176
 
193
- ## How It Works
177
+ | Action | Description |
178
+ |--------|-------------|
179
+ | `state` | Current emotional state (valence, arousal, trend, sensitivity) |
180
+ | `signals` | Recent emotional signals with strength and source |
181
+ | `timeline` | Historical emotional data |
182
+ | `kg_weights` | Entities weighted by emotional significance |
183
+ | `spikes` | Detected emotional spikes |
184
+ | `acknowledge` | Mark a spike as acknowledged |
185
+ | `summary` | AI-generated emotional summary |
194
186
 
195
- ```
196
- Your AI Agent ←→ MCP Protocol ←→ MIND MCP Server ←→ MIND Knowledge Graph
197
- ```
187
+ ### `mind_research` — Autonomous Research
198
188
 
199
- 1. Agent calls `mind_context` at session start → gets identity, rules, priorities
200
- 2. Agent calls `mind_query` before decisions → gets relevant memories
201
- 3. Agent calls `mind_remember` after tasks → stores outcomes and learnings
202
- 4. Next session, same agent (or different one!) has full context
189
+ | Action | Description |
190
+ |--------|-------------|
191
+ | `start` | Launch a deep research job on any topic |
192
+ | `status` | Check job progress |
193
+ | `list` | View all research jobs |
194
+
195
+ ### `mind_train` — Self-Training
203
196
 
204
- **Your knowledge graph grows from every tool you use.** What you learn in Claude Code is available in Cursor. What you decide in OpenClaw is remembered by Claude Desktop.
197
+ | Action | Description |
198
+ |--------|-------------|
199
+ | `start` | Begin guided training (basics, network, expertise, history, goals, freeform) |
200
+ | `chat` | Send training message |
201
+ | `status`, `list_sessions`, `pause`, `resume` | Session management |
202
+ | `save_chat` | Save any chat conversation into the knowledge graph |
205
203
 
206
- ## Why MIND vs. Flat File Memory
204
+ ### `mind_social` — Social Layer
207
205
 
208
- | | Flat files (MEMORY.md) | MIND |
209
- |---|---|---|
210
- | Size limit | ~20K chars, then truncated | Unlimited graph |
211
- | Retrieval | Loads everything every turn | Only relevant memories |
212
- | Capture | Manual (agent must write) | Automatic categorization |
213
- | Cross-agent | One tool only | Shared across all agents |
214
- | Structure | Unstructured text | Knowledge graph with entities & relationships |
215
- | Search | Keyword/vector on small file | Hybrid semantic + graph traversal |
206
+ | Action | Description |
207
+ |--------|-------------|
208
+ | `create_thought`, `get_thought`, `delete_thought`, `like_thought` | Thought (post) management |
209
+ | `feed`, `user_feed`, `search_feed` | Social feed browsing |
210
+ | `create_community`, `list_communities`, `get_community` | Community management |
211
+ | `join_community`, `leave_community` | Membership |
212
+ | `create_post`, `list_posts` | Community posts |
213
+
214
+ ### `mind_profile` — Profile & Preferences
215
+
216
+ | Action | Description |
217
+ |--------|-------------|
218
+ | `get`, `update` | Profile management |
219
+ | `get_chat_prompt`, `set_chat_prompt` | Custom chat system prompt |
220
+ | `get_thought_prompt`, `set_thought_prompt` | Custom thought generation prompt |
221
+ | `get_model`, `set_model`, `list_models` | LLM model selection (50+ models) |
222
+
223
+ ### `mind_insights` — Autonomous Learning Engine
224
+
225
+ | Action | Description |
226
+ |--------|-------------|
227
+ | `list` | Recent pattern-detected insights |
228
+ | `unread_count` | Count of unseen insights |
229
+ | `view`, `feedback` | Mark seen, rate helpfulness |
230
+ | `analyze` | Trigger on-demand analysis |
231
+ | `weekly_summary` | Weekly intelligence summary |
232
+ | `context` | ALE context data |
233
+
234
+ ### `mind_automate` — Automations
235
+
236
+ | Action | Description |
237
+ |--------|-------------|
238
+ | `list`, `create`, `update`, `delete` | Automation CRUD |
239
+ | `run_now` | Trigger immediately |
240
+ | `history` | Execution log |
241
+
242
+ ### `mind_notify` — Notifications
243
+
244
+ | Action | Description |
245
+ |--------|-------------|
246
+ | `list` | View all notifications |
247
+ | `mark_read`, `mark_all_read` | Read management |
248
+ | `stats` | Notification overview |
249
+
250
+ ### `mind_admin` — Administration
251
+
252
+ | Action | Description |
253
+ |--------|-------------|
254
+ | `create_user` | Provision new MIND account |
255
+ | `list_users` | List users with analytics |
256
+ | `update_user_tier` | Change subscription tier |
257
+ | `adjust_user_credits` | Add/deduct credits |
258
+ | `create_featured_mind` | Create public featured mind |
259
+ | `list_featured_minds`, `update_featured_mind` | Featured minds catalog |
260
+
261
+ ## Partner Integration
262
+
263
+ Partner apps can programmatically create MIND accounts using partner keys:
264
+
265
+ ```bash
266
+ curl -X POST https://m-i-n-d.ai/admin/users/create \
267
+ -H "X-API-Key: mind_partner_YOUR_KEY" \
268
+ -H "Content-Type: application/json" \
269
+ -d '{
270
+ "username": "newuser",
271
+ "email": "user@app.com",
272
+ "password": "securepassword",
273
+ "generate_api_key": true
274
+ }'
275
+ ```
276
+
277
+ Returns a JWT + permanent API key. Your app stores the API key and uses it for all subsequent MCP/API calls on behalf of that user.
216
278
 
217
279
  ## Environment Variables
218
280
 
@@ -235,6 +297,17 @@ const server = createMindMcpServer(client);
235
297
  // Connect to any MCP transport
236
298
  ```
237
299
 
300
+ ## Patents
301
+
302
+ MIND's technology is protected by multiple provisional patents including:
303
+ - Emotion-Weighted Knowledge Graph Encoding (U.S. App. 64/030,662)
304
+ - Cross-Agent Persistent Memory via Model Context Protocol
305
+
306
+ ## Links
307
+
308
+ - **Website**: [m-i-n-d.ai](https://m-i-n-d.ai)
309
+ - **npm**: [@astramindapp/mcp-server](https://www.npmjs.com/package/@astramindapp/mcp-server)
310
+
238
311
  ## License
239
312
 
240
- MIT — Astra AI
313
+ MIT — Astrai AI, Inc.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@astramindapp/mcp-server",
3
- "version": "0.5.0",
3
+ "version": "0.5.1",
4
4
  "description": "MIND MCP Server — Universal AI memory layer. Give any LLM agent persistent memory, structured context, and automatic knowledge capture via the Model Context Protocol.",
5
5
  "files": [
6
6
  "dist",