OpenWork ID

Open standard for portable, peer-verified professional identity.

---

What is OpenWork ID?

OpenWork ID defines an open, JSON-LD-based schema for professional identity that is:

• Portable — Export your complete professional identity as structured data. Zero lock-in.
• Peer-verified — Real people publicly vouch for your work, staking their own reputation.
• AI-agent ready — Native MCP (Model Context Protocol) support. Agents read verified data without scraping.
• Free forever — No paywalls, no premium tiers. Professionals pay nothing.

The standard does not own identity. It defines how identity travels between systems.

Website: openworkid.org Reference implementation: upstand.work

---

Core Concepts

The schema defines five interdependent concepts:

1. Experience Record

A structured narrative of professional work — not a CV line item. Each record captures context, challenge, contribution, result, and learning.

{
  "@type": "ExperienceRecord",
  "id": "https://upstand.work/maria/exp/abc123",
  "organisation": "Acme GmbH",
  "role": "Senior Backend Engineer",
  "startDate": "2022-03",
  "endDate": "2024-01",
  "context": "Legacy payment system processing 2M transactions/month...",
  "challenge": "Migration to event-driven architecture without downtime...",
  "contribution": "I designed the dual-write strategy and led a team of 4...",
  "result": "Zero-downtime migration completed in 8 weeks, processing latency reduced by 62%.",
  "learning": "Learned that incremental rollouts with feature flags reduce risk more than big-bang releases."
}

2. Peer Verification

Public confirmation by a named individual. Verifiers stake their professional identity — name, role, and company are visible to everyone.

{
  "@type": "PeerVerification",
  "verifier": {
    "name": "Thomas Weber",
    "role": "CTO",
    "organisation": "Acme GmbH",
    "emailDomain": "acme.de",
    "emailDomainType": "corporate"
  },
  "statement": "Maria led this migration with remarkable clarity. The dual-write approach was her idea and it worked flawlessly.",
  "statementType": "text",
  "mutualFlag": false,
  "withdrawnAt": null
}

3. Human Proof

Evidence that a real person controls the profile — not an AI agent claiming credentials.

Method	Strength	Description
email-magic-link	Minimum	Email ownership verified
peer-verification-chain	Strong	Multiple active peer verifications
document-verified	Strongest	Government ID verification (future)

4. Portable Identity

Complete export envelope for cross-platform transfer. Includes all experience records, verifications, extensions, and export metadata.

5. Extensions

Modular, profession-specific field sets. Each extension has a canonical URI, is independently versioned, and attaches to any ProfessionalIdentity. The core schema stays universal — extensions make a Psychotherapist's profile structurally different from a Software Engineer's.

---

Extension Library

11 extensions across three categories, all at v1.0.0:

Technical

Extension	Key Fields
Software Engineer	languages, frameworks, cloud, databases, seniority, AI tools
Data Scientist	ML frameworks, specialisations, cloud ML, publications
Product Manager	product types, methodologies, team size, tech affinity
UX Designer	disciplines, tools, project types, portfolio

Business

Extension	Key Fields
Finance Controller	specialisations (FP&A, Treasury, M&A...), ERP systems, certifications
Management Consultant	strategy, transformation, engagement length, methodologies
Creative Director	disciplines, industries, project types, portfolio

Regulated

Extension	Key Fields
Psychotherapist	approbation, therapeutic approach, setting, Kassenzulassung
Physician	specialty, subspecialty, setting, certifications
Lawyer	legal areas, bar admission, Fachanwaltstitel
Architect	Leistungsphasen (HOAI), building types, BIM, chamber membership

Regulated extensions include a mandatory regulatoryNotice field. Implementing platforms must display this notice during profile editing.

---

MCP Server

Native Model Context Protocol support. AI agents can read verified professional data without scraping.

Server: https://mcp.upstand.work/sse (MCP 2024-11-05) REST fallback: https://api.upstand.work/v1

Tool	Auth	Status	Description
get_profile	Public	Available	Full professional identity document
get_verifications	Public	Available	Peer verifications for a profile
get_suggestions	User	Available	AI-generated profile improvement suggestions
get_market_benchmark	User	In development	Anonymised rate benchmarks
add_experience	User	In development	Create draft experience record
request_verification	User	In development	Send verification request
search_profiles	Client	Planned	Search verified profiles with filters

See MCP Server Documentation for full tool signatures, auth levels, and rate limits.

---

Versioning

Current version: v0.1.0 (April 2026)

The standard follows semantic versioning:

• MAJOR — Breaking changes (field removals, type changes). Existing documents may fail validation.
• MINOR — Additive changes (new optional fields, enum values, extensions). Existing documents remain valid.
• PATCH — Documentation and example fixes. Zero schema changes.

Roadmap

Version	Milestone	Key Changes
v0.1.0	M1	Initial release. Core schema, 11 extensions, 3 MCP tools
v0.2.0	M2	M2 MCP tools, VerifiedEmployer schema, ContactRequest, RFC process opens
v1.0.0	M3	Stable release. search_profiles, VerifierReputation, Foundation governance

---

Governance

Three Immovable Principles

1. Professionals pay nothing. Ever. No paywalls, no premium tiers.
2. Your data is yours. Always. Export as JSON anytime. Delete within 30 days.
3. The standard belongs to no one. MIT-licensed. Governance transfers to a neutral foundation when adoption reaches critical mass.

RFC Process

All schema changes are discussed and decided publicly on GitHub:

1. Propose — Open a discussion in Ideas describing the problem and your proposed change
2. Discuss — Community feedback and iteration in the discussion thread
3. Submit — Once there's rough consensus, submit a PR using the RFC template
4. Review — Comment period: 7 days (non-breaking) or 30 days (breaking)
5. Merge — Changelog entry written before merge

Releases and breaking changes are announced in Announcements. Questions about the spec belong in Q&A.

---

Repository Structure

openworkid.org/
├── schema/
│   └── v1/
│       ├── professional-identity.jsonld   # Core schema
│       ├── context.jsonld                  # JSON-LD @context
│       └── extensions/                     # 11 profession-specific extensions
│           ├── software-engineer.json
│           ├── data-scientist.json
│           ├── product-manager.json
│           ├── ux-designer.json
│           ├── finance-controller.json
│           ├── management-consultant.json
│           ├── creative-director.json
│           ├── psychotherapist.json
│           ├── physician.json
│           ├── lawyer.json
│           └── architect.json
├── docs/
│   └── mcp-server.md                      # MCP server specification
├── rfcs/
│   └── TEMPLATE.md                        # RFC template for schema changes
├── CHANGELOG.md
├── CONTRIBUTING.md
├── CODE_OF_CONDUCT.md
└── LICENSE

---

Quick Start

Read a profile (MCP)

# Using any MCP-compatible client
curl -X POST https://mcp.upstand.work/sse \
  -H "Content-Type: application/json" \
  -d '{"method": "tools/call", "params": {"name": "get_profile", "arguments": {"username": "maria"}}}'

Read a profile (REST)

curl https://api.upstand.work/v1/profile/maria

Export as JSON-LD

curl https://api.upstand.work/v1/export/maria \
  -H "Accept: application/ld+json"

---

Contributing

See CONTRIBUTING.md for how to propose changes, submit extensions, and participate in the RFC process.

---

License

MIT. See LICENSE.

---

Built and maintained by entwicklerherz. Reference implementation at upstand.work.