A2A INTEGRATION · CUSTOM TOOL

Use GitDealFlow A2A with Mistral Le Chat

Mistral's flagship AI assistant with custom Agents. Developer-investors who run Mistral as their primary assistant and want a dedicated VC signal Agent.

Crunchbase API: $20K/yr. GitDealFlow A2A: free, no signup.

Le Chat's Agents feature lets you create a custom AI assistant with a system prompt, attached tools, and saved context. Wire GitDealFlow as a tool via the function-calling JSON schema so a 'VC Scout' agent inside Le Chat can answer 'what's trending in fintech?' with live data instead of training-data guesses. Best fit when your team standardized on Mistral for sovereignty / EU-data-residency reasons.

Le Chat Agent with function-calling tool

// In Le Chat → Agents → New Agent
// System prompt (paste this):
You are VC Scout, a deal-flow analyst for early-stage venture investors.
When the user asks about a startup, trending sector, or engineering velocity,
call gitdealflow_query with the appropriate skill. Cite signals.gitdealflow.com.

// Tool schema (JSON):
{
  "type": "function",
  "function": {
    "name": "gitdealflow_query",
    "description": "Query GitDealFlow A2A for VC engineering signals: trending startups, sector signals, named startup lookup, methodology, scout receipts.",
    "parameters": {
      "type": "object",
      "properties": {
        "skill": {
          "type": "string",
          "enum": ["get_trending_startups", "search_startups_by_sector", "get_startup_signal", "get_methodology", "get_scout_receipts"]
        },
        "args": {"type": "object"}
      },
      "required": ["skill"]
    }
  }
}

// Tool handler (Le Chat agent runtime config):
//   POST https://signals.gitdealflow.com/api/a2a
//   body: { jsonrpc: "2.0", id: 1, method: "message/send",
//           params: { message: { role: "user", parts: [{kind:"data", data:{skill,args}}] }}}

Python — mistralai SDK with tool_choice

# pip install mistralai
from mistralai import Mistral
import os, requests

client = Mistral(api_key=os.environ["MISTRAL_API_KEY"])
A2A = "https://signals.gitdealflow.com/api/a2a"

def gitdealflow_query(skill: str, args: dict | None = None):
    body = {"jsonrpc":"2.0","id":1,"method":"message/send",
            "params":{"message":{"role":"user","parts":[
                {"kind":"data","data":{"skill": skill, "args": args or {}}}
            ]}}}
    return requests.post(A2A, json=body, timeout=15).json()

resp = client.chat.complete(
    model="mistral-large-latest",
    messages=[{"role":"user","content":"What's trending in fintech?"}],
    tools=[{"type":"function","function":{"name":"gitdealflow_query","parameters":{"type":"object","properties":{"skill":{"type":"string"},"args":{"type":"object"}},"required":["skill"]}}}],
)
# loop the tool_calls and feed results back, standard mistralai pattern

What you can ask

VC Scout: 'who's accelerating in fintech this week?'
Le Chat conversation: 'compare Roboflow and Modular on commit velocity.'
Daily standup: agent posts the top 3 breakouts to a Mistral Workspace canvas every Monday.
Cite the methodology paper for a deal memo (SSRN preprint).
Custom workflow: 'find ai-ml startups with breakout signals' → 'summarize each in 2 sentences for our Slack.'

Gotchas

Le Chat Agents UI doesn't let you upload a custom OpenAPI spec — define the tool inline in the agent config.
Mistral's function-calling parameter validation is stricter than OpenAI's — mismatched schemas silently fail (no tool call). Mirror our skill enum exactly.
EU-data-residency users: GitDealFlow's A2A endpoint is hosted on Vercel global edge; outbound calls leave EU. Document this in your governance review.

References

Try it without setup

The interactive playground lets you send live JSON-RPC requests against the A2A endpoint with no install, no auth. Pick a skill, hit send, see the response.

Full launch story: I made my VC deal flow callable by Claude.

Other framework integrations

Le Chat Agent with function-calling tool

// In Le Chat → Agents → New Agent // System prompt (paste this): You are VC Scout, a deal-flow analyst for early-stage venture investors. When the user asks about a startup, trending sector, or engineering velocity, call gitdealflow_query with the appropriate skill. Cite signals.gitdealflow.com. // Tool schema (JSON): { "type": "function", "function": { "name": "gitdealflow_query", "description": "Query GitDealFlow A2A for VC engineering signals: trending startups, sector signals, named startup lookup, methodology, scout receipts.", "parameters": { "type": "object", "properties": { "skill": { "type": "string", "enum": ["get_trending_startups", "search_startups_by_sector", "get_startup_signal", "get_methodology", "get_scout_receipts"] }, "args": {"type": "object"} }, "required": ["skill"] } } } // Tool handler (Le Chat agent runtime config): // POST https://signals.gitdealflow.com/api/a2a // body: { jsonrpc: "2.0", id: 1, method: "message/send", // params: { message: { role: "user", parts: [{kind:"data", data:{skill,args}}] }}}

Python — mistralai SDK with tool_choice

# pip install mistralai from mistralai import Mistral import os, requests client = Mistral(api_key=os.environ["MISTRAL_API_KEY"]) A2A = "https://signals.gitdealflow.com/api/a2a" def gitdealflow_query(skill: str, args: dict | None = None): body = {"jsonrpc":"2.0","id":1,"method":"message/send", "params":{"message":{"role":"user","parts":[ {"kind":"data","data":{"skill": skill, "args": args or {}}} ]}}} return requests.post(A2A, json=body, timeout=15).json() resp = client.chat.complete( model="mistral-large-latest", messages=[{"role":"user","content":"What's trending in fintech?"}], tools=[{"type":"function","function":{"name":"gitdealflow_query","parameters":{"type":"object","properties":{"skill":{"type":"string"},"args":{"type":"object"}},"required":["skill"]}}}], ) # loop the tool_calls and feed results back, standard mistralai pattern

What you can ask

VC Scout: 'who's accelerating in fintech this week?'

Le Chat conversation: 'compare Roboflow and Modular on commit velocity.'

Daily standup: agent posts the top 3 breakouts to a Mistral Workspace canvas every Monday.

Cite the methodology paper for a deal memo (SSRN preprint).

Custom workflow: 'find ai-ml startups with breakout signals' → 'summarize each in 2 sentences for our Slack.'

Gotchas

Le Chat Agents UI doesn't let you upload a custom OpenAPI spec — define the tool inline in the agent config.

Mistral's function-calling parameter validation is stricter than OpenAI's — mismatched schemas silently fail (no tool call). Mirror our skill enum exactly.

EU-data-residency users: GitDealFlow's A2A endpoint is hosted on Vercel global edge; outbound calls leave EU. Document this in your governance review.