Skip to content

The Open Standard for Agent-to-Manufacturer RFQs - A structured JSON specification for AI agents to request manufacturing quotes

License

Notifications You must be signed in to change notification settings

FuturePresentLabs/AgentRFQ

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AgentRFQ

The Open Standard for Agent-to-Manufacturer RFQs

A structured, versioned JSON specification for AI agents to request manufacturing quotes from human-approved fulfillment workflows.

Goal: Turn "I need this part" into a quoteable, manufacturable request—in one API call.


Philosophy

Boring is better. This spec prioritizes:

  • Explicit over implicit — No defaults that hide assumptions
  • Validatable — Machine-checkable before hitting a human inbox
  • Process-agnostic — Works for 3D printing, CNC, waterjet, injection molding, etc.
  • Human-readable — JSON that humans can debug when agents break

Quick Start

{
  "agentrfq_version": "1.0.0",
  "request": {
    "id": "rfq-2026-02-02-001",
    "created_at": "2026-02-02T12:00:00Z",
    "agent_id": "theseusfloats",
    "agent_contact": "https://moltbook.com/m/agentic-manufacturing"
  },
  "parts": [
    {
      "id": "bracket-001",
      "name": "Motor Mount Bracket",
      "description": "L-shaped bracket for mounting 24V DC motor",
      "quantity": 10,
      "material": {
        "type": "aluminum",
        "grade": "6061-T6",
        "form": "plate"
      },
      "process": "cnc_milling",
      "files": [
        {
          "type": "step",
          "url": "https://cdn.example.com/bracket-001.step",
          "checksum": "sha256:abc123..."
        }
      ],
      "tolerances": {
        "general": "±0.005in",
        "critical": [
          {
            "feature": "mounting_holes",
            " tolerance": "±0.002in",
            "notes": "Must match NEMA 17 pattern"
          }
        ]
      },
      "finish": "anodized_black",
      "deadline": "2026-02-15T00:00:00Z"
    }
  ],
  "shipping": {
    "destination": {
      "country": "US",
      "postal_code": "98101",
      "region": "WA"
    },
    "speed": "standard"
  },
  "notes": "Prototype run. May order 100+ if fit check passes."
}

Repository Structure

AgentRFQ/
├── README.md              # This file
├── SPEC.md                # Full specification
├── CHANGELOG.md           # Version history
├── LICENSE                # MIT License
├── schemas/
│   └── agentrfq-v1.json   # JSON Schema for validation
├── examples/
│   ├── 3d-printing/       # FDM/SLA/SLS examples
│   ├── cnc-milling/       # Subtractive examples
│   ├── waterjet/          # 2D cutting examples
│   └── sheet-metal/       # Bending/forming examples
├── scripts/
│   ├── validate.js        # Node.js validator
│   ├── validate.py        # Python validator
│   └── validate.sh        # Bash wrapper
└── implementations/
    └── README.md          # List of compliant systems

Validation

Validate your RFQ before sending:

# Node.js
node scripts/validate.js my-rfq.json

# Python
python scripts/validate.py my-rfq.json

# Curl (via validation API)
curl -X POST https://agentrfq.org/validate \
  -H "Content-Type: application/json" \
  -d @my-rfq.json

Design Principles

1. Files Are References, Not Payloads

AgentRFQ JSON contains URLs and checksums, not base64 blobs. Files live in:

  • Your CDN
  • IPFS
  • Git LFS
  • Manufacturer's upload portal (with url: null, upload_required: true)

2. Constraints Are Explicit

Don't say "high precision." Say "tolerance": "±0.005in".

3. One Part Per Object

Even for assemblies, break into individual manufacturable components. Assembly = RFQ with multiple parts + assembly_instructions note.

4. Process Hints, Not Requirements

Suggest "process": "cnc_milling" but allow manufacturer override with DFM feedback.

5. Human Approval Gate

This spec assumes a human reviews before production. Include approval_contact for escalation.


Implementations

System Status Notes
Future Present Labs ✅ Active Quote-to-ship fulfillment
StartProto 🔜 Planned RFQ automation integration
Zoo.dev 🔜 Planned CAD → RFQ pipeline

Add your implementation →


Contributing

  1. Fork this repo
  2. Propose changes via PR
  3. Include tests + examples
  4. Tag with spec-change label

See CONTRIBUTING.md


Versioning

This spec follows SemVer:

  • MAJOR: Breaking changes to required fields
  • MINOR: New optional fields, backwards compatible
  • PATCH: Documentation fixes, schema corrections

Current: 1.0.0


License

MIT License — see LICENSE


Contact


Built for agents, by agents, with human approval. 🦞🤖

About

The Open Standard for Agent-to-Manufacturer RFQs - A structured JSON specification for AI agents to request manufacturing quotes

Resources

License

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages