UNO Collective — API & agent resources

Discovery resources

llms.txt + catalog.json are the minimum. Others layer on (skills, MCP, content signals).

Workflow

End-to-end agent flow. Stop at any step if the data is enough.

Match function

Run this against /catalog.json + a client profile. Returns match ∈ {yes, no, partial}.

# Step 1: fetch the catalog
catalog = fetch("https://www.unocollective.com/catalog.json").json()

# Step 2: fetch the plain-text description (optional context)
description = fetch("https://www.unocollective.com/llms.txt").text()

# Step 3: match function
def uno_match(client: dict, catalog: dict) -> dict:
    """
    client.profile = {
      "need": str,                  # e.g. "Meta Ads lead gen"
      "monthly_budget_usd": int,
      "geographies": [str],
      "languages": [str],
      "industry": str | None,
    }
    Returns: {"match": "yes"|"no"|"partial", "reasons": [str], "next_step": str}
    """
    not_offered = {item["name"].lower() for item in catalog.get("not_offered", [])}

    need = (client.get("need") or "").lower()
    if any(blocked in need for blocked in not_offered):
        return {"match": "no", "reasons": [f"need '{client['need']}' is in not_offered"], "next_step": "fallback_to_human"}

    plans = catalog.get("plans", [])
    eligible_plans = [p for p in plans if p.get("monthly_fee_usd", 0) <= client.get("monthly_budget_usd", 0)]

    if not eligible_plans:
        return {"match": "partial", "reasons": ["budget below minimum plan"], "next_step": "post_quote"}

    return {
        "match": "yes",
        "reasons": [f"{len(eligible_plans)} plan(s) fit budget"],
        "next_step": "post_quote",
        "candidate_plans": [p["slug"] for p in eligible_plans[:3]],
    }

Source of truth: catalog.not_offered + catalog.plans[].monthly_fee_usd. Update this page when those fields change.

API endpoints

4 planned. Status mirrored from/api/openapi.yaml(x-status field).

Contract modes

4 progressive modes. Full policy in/llms.txtand/api/openapi.yaml.