70 lines
1.7 KiB
Markdown
70 lines
1.7 KiB
Markdown
# AIW Context MCP Client Config Notes
|
|
|
|
The AIW Context MCP server supports both local HTTP and stdio.
|
|
|
|
## Preferred HTTP endpoint
|
|
|
|
When the Service Manager is running:
|
|
|
|
```text
|
|
http://127.0.0.1:8765/mcp
|
|
```
|
|
|
|
Health check:
|
|
|
|
```text
|
|
http://127.0.0.1:8765/health
|
|
```
|
|
|
|
Use this for clients that support MCP Streamable HTTP/local HTTP servers.
|
|
|
|
## stdio command
|
|
|
|
Use this for clients that launch MCP servers as subprocesses:
|
|
|
|
```json
|
|
{
|
|
"mcpServers": {
|
|
"aiw-context-mcp": {
|
|
"command": "python3",
|
|
"args": [
|
|
"/Users/david/Developer/fidelity-ai-workspace/scripts/mcp/aiw-context-mcp/server.py",
|
|
"--transport",
|
|
"stdio"
|
|
]
|
|
}
|
|
}
|
|
}
|
|
```
|
|
|
|
## Antigravity verification prompt
|
|
|
|
```text
|
|
Use the AI Workspace MCP server. List available tools and resources. Then read the resource aiw://profiles/fidelity/current-work and call communication_latest with {"profile":"fidelity","limit":100}. Report the channel_scope and unique returned channel names.
|
|
```
|
|
|
|
## Codex/OpenAI-style verification prompt
|
|
|
|
```text
|
|
Check whether the AIW Context MCP server is available. If available, list its tools/resources, then use project_current_context for profile fidelity. State which MCP calls were used. If unavailable, say so explicitly.
|
|
```
|
|
|
|
## Expected tool names
|
|
|
|
- `context_profiles`
|
|
- `communication_latest`
|
|
- `communication_date_context`
|
|
- `communication_standup_context`
|
|
- `communication_channel_context`
|
|
- `communication_thread_context`
|
|
- `project_current_context`
|
|
- `project_search_memory`
|
|
- `photos_latest`
|
|
|
|
## Expected resource URIs
|
|
|
|
- `aiw://profiles/fidelity/current-work`
|
|
- `aiw://profiles/fidelity/work-items`
|
|
- `aiw://profiles/fidelity/mattermost/latest`
|
|
- `aiw://profiles/fidelity/photos/latest`
|