cortexdb-mcp 0.2.2__tar.gz → 0.3.1__tar.gz

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.
@@ -0,0 +1,79 @@
1
+ # Rust
2
+ /target
3
+ **/*.rs.bk
4
+
5
+ # Environment / secrets
6
+ .env
7
+ .env.local
8
+ .env*.local
9
+ *.pem
10
+ *.key
11
+ .npmrc
12
+
13
+ # SQLite database
14
+ *.sqlite
15
+ *.sqlite-wal
16
+ *.sqlite-shm
17
+
18
+ # OS
19
+ .DS_Store
20
+ Thumbs.db
21
+ desktop.ini
22
+
23
+ # IDE
24
+ .idea/
25
+ .vscode/
26
+ *.swp
27
+ *.swo
28
+
29
+ # Data directories
30
+ cortexdb_data*/
31
+ /data/
32
+ # Per-bench tenant stores (RocksDB + Tantivy + HNSW state; regeneratable per run)
33
+ /data_*/
34
+ # Experimental per-branch stores (not tracked on this branch but left gitignored
35
+ # so checkout from other branches doesn't surface them in git status)
36
+ /event_memory_store/
37
+ /llm_cache/
38
+
39
+ # Benchmark inputs and per-run outputs (kept local, regenerated each run)
40
+ benchmarks/longmemeval/data/
41
+ benchmarks/longmemeval/server_results/
42
+ benchmarks/longmemeval/fast_results/
43
+ benchmarks/longmemeval/micro_results/
44
+ benchmarks/longmemeval/server_logs/
45
+ benchmarks/longmemeval/*.log
46
+ benchmarks/locomo/locomo_results*.json
47
+ benchmarks/locomo/server_results/
48
+ benchmarks/locomo/*.log
49
+ /answer_out.json
50
+
51
+ # Local Claude Code state
52
+ .claude/
53
+ .tmp/
54
+
55
+ # Python
56
+ __pycache__/
57
+ *.pyc
58
+ .venv/
59
+ venv/
60
+
61
+ # Node
62
+ node_modules/
63
+ dist/
64
+ .next/
65
+
66
+ # Egg info
67
+ *.egg-info/
68
+
69
+ # Scratch/debug text files at root
70
+ /*.txt
71
+ /*.log
72
+
73
+ # Local debug / marketing / private content (not for repo)
74
+ harness/.reports/
75
+ harness_data_*/
76
+ blog/
77
+ sales/
78
+ videos/
79
+ local-instance/
@@ -1,12 +1,12 @@
1
- FROM python:3.12-slim
2
-
3
- WORKDIR /app
4
-
5
- COPY pyproject.toml README.md ./
6
- COPY cortexdb_mcp/ cortexdb_mcp/
7
-
8
- RUN pip install --no-cache-dir .
9
-
10
- ENV CORTEXDB_URL=https://api.cortexdb.ai
11
-
12
- CMD ["cortexdb-mcp"]
1
+ FROM python:3.12-slim
2
+
3
+ WORKDIR /app
4
+
5
+ COPY pyproject.toml README.md ./
6
+ COPY cortexdb_mcp/ cortexdb_mcp/
7
+
8
+ RUN pip install --no-cache-dir .
9
+
10
+ ENV CORTEXDB_URL=https://api-v1.cortexdb.ai
11
+
12
+ CMD ["cortexdb-mcp"]
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: cortexdb-mcp
3
- Version: 0.2.2
3
+ Version: 0.3.1
4
4
  Summary: MCP Server for CortexDB — expose memory operations to AI agents
5
5
  License-Expression: MIT
6
6
  Requires-Python: >=3.10
@@ -15,31 +15,33 @@ Description-Content-Type: text/markdown
15
15
 
16
16
  MCP (Model Context Protocol) server that gives AI tools persistent long-term memory via CortexDB. Works with Claude Desktop, Cursor, Windsurf, VS Code Copilot, and any MCP-compatible client.
17
17
 
18
+ > **0.3.0 rewrites the server against the v1 CortexDB API.** Default surface is now `https://api-v1.cortexdb.ai`. The server **auto-signs-up anonymously** on first launch — no API key required.
19
+
18
20
  ## Quick Start
19
21
 
20
22
  ```bash
21
- # Install
22
- pip install -e .
23
+ # Install (the PyPI name is `cortexdb-mcp`)
24
+ pip install cortexdb-mcp
23
25
 
24
- # Or with uvx (no install needed)
25
- uvx --from . cortexdb-mcp
26
+ # Then run with zero config:
27
+ cortexdb-mcp
26
28
  ```
27
29
 
28
- Set your environment variables:
30
+ That's it. On first launch the server hits `POST /v1/auth/signup`, mints a free-tier PASETO token + scope for itself, and caches them under `~/.config/cortexdb-mcp/state.json` (Linux/macOS) or `%APPDATA%\cortexdb-mcp\state.json` (Windows). Re-launches reuse the cached identity until the token expires (7 days).
29
31
 
30
- ```bash
31
- export CORTEXDB_URL="https://api.cortexdb.ai"
32
- export CORTEXDB_API_KEY="cx_live_your_key_here"
33
- ```
34
-
35
- Run:
32
+ To target a custom deployment or pre-existing identity, set any of:
36
33
 
37
34
  ```bash
38
- cortexdb-mcp
35
+ export CORTEXDB_URL="https://api-v1.cortexdb.ai" # base URL
36
+ export CORTEXDB_API_KEY="v4.public..." # PASETO bearer token
37
+ export CORTEXDB_ACTOR="user:alice" # ActorId, sent as X-Cortex-Actor
38
+ export CORTEXDB_SCOPE="org:acme/user:alice" # default scope path
39
39
  ```
40
40
 
41
41
  ## IDE Setup
42
42
 
43
+ Most clients just need a single line — the auto-signup handles the rest.
44
+
43
45
  ### Claude Desktop
44
46
 
45
47
  Edit `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `%APPDATA%\Claude\claude_desktop_config.json` (Windows):
@@ -48,16 +50,34 @@ Edit `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) o
48
50
  {
49
51
  "mcpServers": {
50
52
  "cortexdb": {
51
- "command": "cortexdb-mcp",
52
- "env": {
53
- "CORTEXDB_URL": "https://api.cortexdb.ai",
54
- "CORTEXDB_API_KEY": "cx_live_your_key_here"
55
- }
53
+ "command": "cortexdb-mcp"
56
54
  }
57
55
  }
58
56
  }
59
57
  ```
60
58
 
59
+ > **Windows note (audit POL-3):** if Claude Desktop / Cursor / Windsurf
60
+ > can't find `cortexdb-mcp` on PATH, give it the full path to the
61
+ > installed executable:
62
+ >
63
+ > ```json
64
+ > {
65
+ > "mcpServers": {
66
+ > "cortexdb": {
67
+ > "command": "C:\\Python313\\Scripts\\cortexdb-mcp.exe"
68
+ > }
69
+ > }
70
+ > }
71
+ > ```
72
+ >
73
+ > Locate it with `where cortexdb-mcp` in `cmd.exe` or
74
+ > `(Get-Command cortexdb-mcp).Source` in PowerShell.
75
+ >
76
+ > Some Windows terminals don't have UTF-8 enabled by default, which can
77
+ > garble emoji in `--help` output (they show as `?`). The server itself
78
+ > isn't affected — only the CLI banner. Run
79
+ > `chcp 65001` once per terminal session to fix.
80
+
61
81
  ### Claude Code (CLI)
62
82
 
63
83
  Edit `~/.claude/mcp.json`:
@@ -66,11 +86,7 @@ Edit `~/.claude/mcp.json`:
66
86
  {
67
87
  "mcpServers": {
68
88
  "cortexdb": {
69
- "command": "cortexdb-mcp",
70
- "env": {
71
- "CORTEXDB_URL": "https://api.cortexdb.ai",
72
- "CORTEXDB_API_KEY": "cx_live_your_key_here"
73
- }
89
+ "command": "cortexdb-mcp"
74
90
  }
75
91
  }
76
92
  }
@@ -84,11 +100,7 @@ Edit `~/.cursor/mcp.json`:
84
100
  {
85
101
  "mcpServers": {
86
102
  "cortexdb": {
87
- "command": "cortexdb-mcp",
88
- "env": {
89
- "CORTEXDB_URL": "https://api.cortexdb.ai",
90
- "CORTEXDB_API_KEY": "cx_live_your_key_here"
91
- }
103
+ "command": "cortexdb-mcp"
92
104
  }
93
105
  }
94
106
  }
@@ -104,11 +116,7 @@ Edit `~/.codeium/windsurf/mcp_config.json`:
104
116
  {
105
117
  "mcpServers": {
106
118
  "cortexdb": {
107
- "command": "cortexdb-mcp",
108
- "env": {
109
- "CORTEXDB_URL": "https://api.cortexdb.ai",
110
- "CORTEXDB_API_KEY": "cx_live_your_key_here"
111
- }
119
+ "command": "cortexdb-mcp"
112
120
  }
113
121
  }
114
122
  }
@@ -122,11 +130,21 @@ Add to `.vscode/mcp.json` in your project or `~/.vscode/mcp.json` globally:
122
130
  {
123
131
  "servers": {
124
132
  "cortexdb": {
125
- "command": "cortexdb-mcp",
126
- "env": {
127
- "CORTEXDB_URL": "https://api.cortexdb.ai",
128
- "CORTEXDB_API_KEY": "cx_live_your_key_here"
129
- }
133
+ "command": "cortexdb-mcp"
134
+ }
135
+ }
136
+ }
137
+ ```
138
+
139
+ ### Windows tip
140
+
141
+ On Windows, MCP clients sometimes need the absolute path:
142
+
143
+ ```json
144
+ {
145
+ "mcpServers": {
146
+ "cortexdb": {
147
+ "command": "C:\\Python310\\Scripts\\cortexdb-mcp.exe"
130
148
  }
131
149
  }
132
150
  }
@@ -136,43 +154,43 @@ Add to `.vscode/mcp.json` in your project or `~/.vscode/mcp.json` globally:
136
154
 
137
155
  ### Memory Operations
138
156
 
139
- | Tool | Description |
140
- |------|-------------|
141
- | `memory_store` | Store a new memory with optional source, type, tags, TTL |
142
- | `memory_search` | Search memories using natural language (hybrid retrieval) |
143
- | `memory_forget` | Delete memories with audit trail (GDPR-compliant) |
144
- | `get_context` | Deep contextual retrieval combining search + graph |
145
- | `advanced_search` | Search with structured filters (source, type, time range) |
146
-
147
- ### Episode Management
148
-
149
- | Tool | Description |
150
- |------|-------------|
151
- | `memory_list` | List episodes with pagination and type filter |
152
- | `memory_get` | Get a specific episode by ID |
153
- | `memory_update` | Update episode content or metadata |
154
- | `memory_delete` | Delete a specific episode by ID |
155
- | `memory_bulk_delete` | Bulk delete with query matching and dry-run support |
156
-
157
- ### Knowledge Graph
158
-
159
- | Tool | Description |
160
- |------|-------------|
161
- | `entity_list` | List entities (people, services, projects, concepts) |
162
- | `entity_get` | Get entity details including relationships and recent episodes |
163
- | `entity_edges` | Get all relationships for an entity |
164
- | `entity_link` | Create a relationship between two entities |
157
+ | Tool | Maps to | Description |
158
+ |------|---------|-------------|
159
+ | `memory_store` | `POST /v1/experience` | Store a new memory. Source/tags/type become labels. |
160
+ | `memory_search` | `POST /v1/recall` | Search memories using natural language. |
161
+ | `memory_forget` | `POST /v1/forget` | Delete memories. With `query`, narrows by subject. |
162
+ | `get_context` | `POST /v1/recall` (holistic) | Deep context with facts + beliefs. |
163
+ | `advanced_search` | `POST /v1/recall` + temporal | Search with structured filters (time / source / type). |
164
+
165
+ ### Event CRUD
166
+
167
+ | Tool | Maps to | Description |
168
+ |------|---------|-------------|
169
+ | `memory_list` | `GET /v1/events` | List events in scope, paginated. |
170
+ | `memory_get` | `GET /v1/events/{id}` | Fetch a single event (used for citations). |
171
+ | `memory_delete` | `POST /v1/forget` (memory_ids) | Delete one event by id. |
172
+ | `memory_bulk_delete` | `POST /v1/forget` or `/v1/erasures/preview` | Bulk delete with dry-run support. |
173
+
174
+ ### Knowledge Graph (facts-backed in v1)
175
+
176
+ | Tool | Maps to | Description |
177
+ |------|---------|-------------|
178
+ | `entity_list` | `GET /v1/facts` | List fact subjects, ranked by count. |
179
+ | `entity_get` | `GET /v1/facts?subject=…` | Fact lineage for a subject. |
180
+ | `entity_edges` | `GET /v1/facts?subject=…` | Predicate/object pairs for an entity. |
181
+ | `entity_link` | `POST /v1/experience` | Store a sentence the extractor will turn into a fact. |
165
182
 
166
183
  ### Admin & Observability
167
184
 
168
- | Tool | Description |
169
- |------|-------------|
170
- | `health_check` | Check CortexDB server health |
171
- | `get_usage` | View usage stats and tier limits |
172
- | `get_insights` | Generate proactive insights (incident spikes, gaps, risks) |
173
- | `get_ontology` | View entity types and relationship types |
174
- | `export_data` | Export memories as JSON |
175
- | `import_data` | Import memories from JSON |
185
+ | Tool | Maps to | Description |
186
+ |------|---------|-------------|
187
+ | `health_check` | `GET /v1/auth/whoami` | Verify the bearer + reach the deployment policy. |
188
+ | `get_usage` | `GET /v1/auth/whoami` + headers | Tier, rate limit, token expiry. |
189
+ | `get_insights` | (in-process) | Proactive insights derived from stored episodes. |
190
+
191
+ ### Removed from 0.3.0
192
+
193
+ `memory_update`, `export_data`, `import_data`, `get_ontology` — v1 has no direct equivalents. Updates: store a new event; bulk import: see `/v1/import` directly with the v1 envelope shape; ontology: not part of the v1 surface.
176
194
 
177
195
  ## Resources
178
196
 
@@ -204,7 +222,7 @@ Pre-built prompt templates:
204
222
 
205
223
  | Environment Variable | Default | Description |
206
224
  |---|---|---|
207
- | `CORTEXDB_URL` | `https://api.cortexdb.ai` | CortexDB server URL |
225
+ | `CORTEXDB_URL` | `https://api-v1.cortexdb.ai` | CortexDB server URL |
208
226
  | `CORTEXDB_API_KEY` | (none) | API key for authentication |
209
227
  | `CORTEXDB_TIMEOUT` | `30.0` | HTTP request timeout (seconds) |
210
228
 
@@ -0,0 +1,261 @@
1
+ # CortexDB MCP Server
2
+
3
+ MCP (Model Context Protocol) server that gives AI tools persistent long-term memory via CortexDB. Works with Claude Desktop, Cursor, Windsurf, VS Code Copilot, and any MCP-compatible client.
4
+
5
+ > **0.3.0 rewrites the server against the v1 CortexDB API.** Default surface is now `https://api-v1.cortexdb.ai`. The server **auto-signs-up anonymously** on first launch — no API key required.
6
+
7
+ ## Quick Start
8
+
9
+ ```bash
10
+ # Install (the PyPI name is `cortexdb-mcp`)
11
+ pip install cortexdb-mcp
12
+
13
+ # Then run with zero config:
14
+ cortexdb-mcp
15
+ ```
16
+
17
+ That's it. On first launch the server hits `POST /v1/auth/signup`, mints a free-tier PASETO token + scope for itself, and caches them under `~/.config/cortexdb-mcp/state.json` (Linux/macOS) or `%APPDATA%\cortexdb-mcp\state.json` (Windows). Re-launches reuse the cached identity until the token expires (7 days).
18
+
19
+ To target a custom deployment or pre-existing identity, set any of:
20
+
21
+ ```bash
22
+ export CORTEXDB_URL="https://api-v1.cortexdb.ai" # base URL
23
+ export CORTEXDB_API_KEY="v4.public..." # PASETO bearer token
24
+ export CORTEXDB_ACTOR="user:alice" # ActorId, sent as X-Cortex-Actor
25
+ export CORTEXDB_SCOPE="org:acme/user:alice" # default scope path
26
+ ```
27
+
28
+ ## IDE Setup
29
+
30
+ Most clients just need a single line — the auto-signup handles the rest.
31
+
32
+ ### Claude Desktop
33
+
34
+ Edit `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `%APPDATA%\Claude\claude_desktop_config.json` (Windows):
35
+
36
+ ```json
37
+ {
38
+ "mcpServers": {
39
+ "cortexdb": {
40
+ "command": "cortexdb-mcp"
41
+ }
42
+ }
43
+ }
44
+ ```
45
+
46
+ > **Windows note (audit POL-3):** if Claude Desktop / Cursor / Windsurf
47
+ > can't find `cortexdb-mcp` on PATH, give it the full path to the
48
+ > installed executable:
49
+ >
50
+ > ```json
51
+ > {
52
+ > "mcpServers": {
53
+ > "cortexdb": {
54
+ > "command": "C:\\Python313\\Scripts\\cortexdb-mcp.exe"
55
+ > }
56
+ > }
57
+ > }
58
+ > ```
59
+ >
60
+ > Locate it with `where cortexdb-mcp` in `cmd.exe` or
61
+ > `(Get-Command cortexdb-mcp).Source` in PowerShell.
62
+ >
63
+ > Some Windows terminals don't have UTF-8 enabled by default, which can
64
+ > garble emoji in `--help` output (they show as `?`). The server itself
65
+ > isn't affected — only the CLI banner. Run
66
+ > `chcp 65001` once per terminal session to fix.
67
+
68
+ ### Claude Code (CLI)
69
+
70
+ Edit `~/.claude/mcp.json`:
71
+
72
+ ```json
73
+ {
74
+ "mcpServers": {
75
+ "cortexdb": {
76
+ "command": "cortexdb-mcp"
77
+ }
78
+ }
79
+ }
80
+ ```
81
+
82
+ ### Cursor
83
+
84
+ Edit `~/.cursor/mcp.json`:
85
+
86
+ ```json
87
+ {
88
+ "mcpServers": {
89
+ "cortexdb": {
90
+ "command": "cortexdb-mcp"
91
+ }
92
+ }
93
+ }
94
+ ```
95
+
96
+ Or in Cursor Settings > MCP Servers > Add Server.
97
+
98
+ ### Windsurf
99
+
100
+ Edit `~/.codeium/windsurf/mcp_config.json`:
101
+
102
+ ```json
103
+ {
104
+ "mcpServers": {
105
+ "cortexdb": {
106
+ "command": "cortexdb-mcp"
107
+ }
108
+ }
109
+ }
110
+ ```
111
+
112
+ ### VS Code (GitHub Copilot)
113
+
114
+ Add to `.vscode/mcp.json` in your project or `~/.vscode/mcp.json` globally:
115
+
116
+ ```json
117
+ {
118
+ "servers": {
119
+ "cortexdb": {
120
+ "command": "cortexdb-mcp"
121
+ }
122
+ }
123
+ }
124
+ ```
125
+
126
+ ### Windows tip
127
+
128
+ On Windows, MCP clients sometimes need the absolute path:
129
+
130
+ ```json
131
+ {
132
+ "mcpServers": {
133
+ "cortexdb": {
134
+ "command": "C:\\Python310\\Scripts\\cortexdb-mcp.exe"
135
+ }
136
+ }
137
+ }
138
+ ```
139
+
140
+ ## Tools
141
+
142
+ ### Memory Operations
143
+
144
+ | Tool | Maps to | Description |
145
+ |------|---------|-------------|
146
+ | `memory_store` | `POST /v1/experience` | Store a new memory. Source/tags/type become labels. |
147
+ | `memory_search` | `POST /v1/recall` | Search memories using natural language. |
148
+ | `memory_forget` | `POST /v1/forget` | Delete memories. With `query`, narrows by subject. |
149
+ | `get_context` | `POST /v1/recall` (holistic) | Deep context with facts + beliefs. |
150
+ | `advanced_search` | `POST /v1/recall` + temporal | Search with structured filters (time / source / type). |
151
+
152
+ ### Event CRUD
153
+
154
+ | Tool | Maps to | Description |
155
+ |------|---------|-------------|
156
+ | `memory_list` | `GET /v1/events` | List events in scope, paginated. |
157
+ | `memory_get` | `GET /v1/events/{id}` | Fetch a single event (used for citations). |
158
+ | `memory_delete` | `POST /v1/forget` (memory_ids) | Delete one event by id. |
159
+ | `memory_bulk_delete` | `POST /v1/forget` or `/v1/erasures/preview` | Bulk delete with dry-run support. |
160
+
161
+ ### Knowledge Graph (facts-backed in v1)
162
+
163
+ | Tool | Maps to | Description |
164
+ |------|---------|-------------|
165
+ | `entity_list` | `GET /v1/facts` | List fact subjects, ranked by count. |
166
+ | `entity_get` | `GET /v1/facts?subject=…` | Fact lineage for a subject. |
167
+ | `entity_edges` | `GET /v1/facts?subject=…` | Predicate/object pairs for an entity. |
168
+ | `entity_link` | `POST /v1/experience` | Store a sentence the extractor will turn into a fact. |
169
+
170
+ ### Admin & Observability
171
+
172
+ | Tool | Maps to | Description |
173
+ |------|---------|-------------|
174
+ | `health_check` | `GET /v1/auth/whoami` | Verify the bearer + reach the deployment policy. |
175
+ | `get_usage` | `GET /v1/auth/whoami` + headers | Tier, rate limit, token expiry. |
176
+ | `get_insights` | (in-process) | Proactive insights derived from stored episodes. |
177
+
178
+ ### Removed from 0.3.0
179
+
180
+ `memory_update`, `export_data`, `import_data`, `get_ontology` — v1 has no direct equivalents. Updates: store a new event; bulk import: see `/v1/import` directly with the v1 envelope shape; ontology: not part of the v1 surface.
181
+
182
+ ## Resources
183
+
184
+ Resources provide read-only data that AI tools can access:
185
+
186
+ | Resource URI | Description |
187
+ |---|---|
188
+ | `cortexdb://health` | Server health status |
189
+ | `cortexdb://metrics` | Request metrics (total, active, errors, rate-limited) |
190
+ | `cortexdb://usage` | Usage statistics and tier limits |
191
+ | `cortexdb://episodes` | Recent 50 episodes |
192
+ | `cortexdb://entities` | Top 100 knowledge graph entities |
193
+ | `cortexdb://insights` | Proactive insights |
194
+ | `cortexdb://ontology` | Entity and relationship type schema |
195
+
196
+ ## Prompts
197
+
198
+ Pre-built prompt templates:
199
+
200
+ | Prompt | Description |
201
+ |---|---|
202
+ | `investigate_incident` | Investigate an incident using stored memories |
203
+ | `summarize_knowledge` | Summarize everything known about a topic |
204
+ | `deployment_review` | Pre-deployment safety review |
205
+ | `onboard_to_codebase` | Onboard to a codebase using stored knowledge |
206
+ | `weekly_digest` | Generate a weekly activity summary |
207
+
208
+ ## Configuration
209
+
210
+ | Environment Variable | Default | Description |
211
+ |---|---|---|
212
+ | `CORTEXDB_URL` | `https://api-v1.cortexdb.ai` | CortexDB server URL |
213
+ | `CORTEXDB_API_KEY` | (none) | API key for authentication |
214
+ | `CORTEXDB_TIMEOUT` | `30.0` | HTTP request timeout (seconds) |
215
+
216
+ ## Examples
217
+
218
+ ### Store a memory from Cursor
219
+
220
+ Ask your AI assistant:
221
+ > "Remember that the payments service was migrated to Stripe v3 on March 15th"
222
+
223
+ The assistant will call `memory_store` with the content.
224
+
225
+ ### Search memories
226
+
227
+ > "What do we know about the payments service?"
228
+
229
+ The assistant calls `memory_search` and gets relevant context from CortexDB.
230
+
231
+ ### Explore the knowledge graph
232
+
233
+ > "Show me all entities related to the auth service"
234
+
235
+ The assistant calls `entity_get` or `entity_edges` to traverse relationships.
236
+
237
+ ### Pre-deployment review
238
+
239
+ > "Run a deployment review for the user-service"
240
+
241
+ Uses the `deployment_review` prompt to check for recent incidents, dependencies, and risks.
242
+
243
+ ## Architecture
244
+
245
+ ```
246
+ ┌─────────────┐ stdio/SSE ┌──────────────┐ HTTP ┌──────────┐
247
+ │ AI Client │ ◄──────────────► │ MCP Server │ ◄──────────► │ CortexDB │
248
+ │ (Cursor, │ MCP JSON-RPC │ (this pkg) │ REST API │ Server │
249
+ │ Claude, │ │ │ │ │
250
+ │ VS Code) │ └──────────────┘ └──────────┘
251
+ └─────────────┘
252
+ ```
253
+
254
+ The MCP server is a thin translation layer:
255
+ 1. Receives MCP tool calls from the AI client
256
+ 2. Translates them to CortexDB REST API calls
257
+ 3. Formats responses for the AI to consume
258
+
259
+ ## License
260
+
261
+ MIT
@@ -1,3 +1,3 @@
1
- """CortexDB MCP Server -- expose CortexDB memory operations to AI agents via MCP."""
2
-
3
- __version__ = "0.2.1"
1
+ """CortexDB MCP Server -- expose CortexDB memory operations to AI agents via MCP."""
2
+
3
+ __version__ = "0.3.1"
@@ -1,5 +1,5 @@
1
- """Allow ``python -m cortexdb_mcp`` to launch the MCP server."""
2
-
3
- from cortexdb_mcp.server import main
4
-
5
- main()
1
+ """Allow ``python -m cortexdb_mcp`` to launch the MCP server."""
2
+
3
+ from cortexdb_mcp.server import main
4
+
5
+ main()