Claude To Medrixflow

MedrixFlow Skill

Communicate with a running MedrixFlow instance via its HTTP API. MedrixFlow is an AI agent platform built on LangGraph that orchestrates sub-agents for research, code execution, web browsing, and more.

Architecture

MedrixFlow exposes two API surfaces behind an Nginx reverse proxy:

Environment Variables

All URLs are configurable via environment variables. Read these env vars before making any request.

When making curl calls, always resolve the URL like this:

# Resolve base URLs from env (do this FIRST before any API call)
MEDRIXFLOW_URL="${MEDRIXFLOW_URL:-http://localhost:2026}"
MEDRIXFLOW_GATEWAY_URL="${MEDRIXFLOW_GATEWAY_URL:-$MEDRIXFLOW_URL}"
MEDRIXFLOW_LANGGRAPH_URL="${MEDRIXFLOW_LANGGRAPH_URL:-$MEDRIXFLOW_URL/api/langgraph}"

Available Operations

1. Health Check

Verify MedrixFlow is running:

curl -s "$MEDRIXFLOW_GATEWAY_URL/health"

2. Send a Message (Streaming)

This is the primary operation. It creates a thread and streams the agent's response.

Step 1: Create a thread

curl -s -X POST "$MEDRIXFLOW_LANGGRAPH_URL/threads" \
  -H "Content-Type: application/json" \
  -d '{}'

Response: {"thread_id": "<uuid>", ...}

Step 2: Stream a run

curl -s -N -X POST "$MEDRIXFLOW_LANGGRAPH_URL/threads/<thread_id>/runs/stream" \
  -H "Content-Type: application/json" \
  -d '{
    "assistant_id": "lead_agent",
    "input": {
      "messages": [
        {
          "type": "human",
          "content": [{"type": "text", "text": "YOUR MESSAGE HERE"}]
        }
      ]
    },
    "stream_mode": ["values", "messages-tuple"],
    "stream_subgraphs": true,
    "config": {
      "recursion_limit": 1000
    },
    "context": {
      "thinking_enabled": true,
      "is_plan_mode": true,
      "subagent_enabled": true,
      "thread_id": "<thread_id>"
    }
  }'

The response is an SSE stream. Each event has the format:

event: <event_type>
data: <json_data>

Key event types:

• metadata — run metadata including run_id
• values — full state snapshot with messages array
• messages-tuple — incremental message updates (AI text chunks, tool calls, tool results)
• end — stream is complete

Context modes (set via context):

• Flash mode: thinking_enabled: false, is_plan_mode: false, subagent_enabled: false
• Standard mode: thinking_enabled: true, is_plan_mode: false, subagent_enabled: false
• Pro mode: thinking_enabled: true, is_plan_mode: true, subagent_enabled: false
• Ultra mode: thinking_enabled: true, is_plan_mode: true, subagent_enabled: true

3. Continue a Conversation

To send follow-up messages, reuse the same thread_id from step 2 and POST another run with the new message.

4. List Models

curl -s "$MEDRIXFLOW_GATEWAY_URL/api/models"

Returns: {"models": [{"name": "...", "provider": "...", ...}, ...]}

5. List Skills

curl -s "$MEDRIXFLOW_GATEWAY_URL/api/skills"

Returns: {"skills": [{"name": "...", "enabled": true, ...}, ...]}

6. Enable/Disable a Skill

curl -s -X PUT "$MEDRIXFLOW_GATEWAY_URL/api/skills/<skill_name>" \
  -H "Content-Type: application/json" \
  -d '{"enabled": true}'

7. List Agents

curl -s "$MEDRIXFLOW_GATEWAY_URL/api/agents"

Returns: {"agents": [{"name": "...", ...}, ...]}

8. Get Memory

curl -s "$MEDRIXFLOW_GATEWAY_URL/api/memory"

Returns user context, facts, and conversation history summaries.

9. Upload Files to a Thread

curl -s -X POST "$MEDRIXFLOW_GATEWAY_URL/api/threads/<thread_id>/uploads" \
  -F "files=@/path/to/file.pdf"

Supports PDF, PPTX, XLSX, DOCX — automatically converts to Markdown.

10. List Uploaded Files

curl -s "$MEDRIXFLOW_GATEWAY_URL/api/threads/<thread_id>/uploads/list"

11. Get Thread History

curl -s "$MEDRIXFLOW_LANGGRAPH_URL/threads/<thread_id>/history"

12. List Threads

curl -s -X POST "$MEDRIXFLOW_LANGGRAPH_URL/threads/search" \
  -H "Content-Type: application/json" \
  -d '{"limit": 20, "sort_by": "updated_at", "sort_order": "desc"}'

Usage Script

For sending messages and collecting the full response, use the helper script:

bash /path/to/skills/claude-to-medrix_flow/scripts/chat.sh "Your question here"

See scripts/chat.sh for the implementation. The script:

1. Checks health
2. Creates a thread
3. Streams the run and collects the final AI response
4. Prints the result

Parsing SSE Output

The stream returns SSE events. To extract the final AI response from a values event:

• Look for the last event: values block
• Parse its data JSON
• The messages array contains all messages; the last one with type: "ai" is the response
• The content field of that message is the AI's text reply

Error Handling

• If health check fails, MedrixFlow is not running. Inform the user they need to start it.
• If the stream returns an error event, extract and display the error message.
• Common issues: port not open, services still starting up, config errors.

Tips

• For quick questions, use flash mode (fastest, no planning).
• For research tasks, use pro or ultra mode (enables planning and sub-agents).
• You can upload files first, then reference them in your message.
• Thread IDs persist — you can return to a conversation later.