shoku

Company: shoku

Self-reported basis: The description is entirely from the author’s own submission to the OpenAI 2026 hackathon on Devpost. It is unverified and contains no evidence of revenue, customers, or traction.

What it appears to be: A mission-based hiring protocol that enables AI agents to hire real people for real-world tasks, using a structured approach to outcomes rather than job titles.

What changed: The project was built as a hackathon submission with the stated goal of exploring how AI agents might interact with human labor in a new kind of marketplace.

Key open question: Is there evidence that this concept has traction or viability beyond a demo-grade prototype?

The description states that shoku is a mission-based hiring protocol where:

• Agents hire real people to get real-world work done.
• A “Mission” is defined as an outcome with boundaries: capability, budget, window, and proof of completion.
• Missions are fractal — small tasks go to individuals; larger ones are broken down into organizational units.
• The system separates planning (LLM) from execution (deterministic code).
• It avoids building a workflow engine by focusing on the hiring unit instead.

Inference: This is not a traditional gig marketplace or task-based platform. It’s a protocol for structuring how AI agents can engage with human labor, using outcomes as the core unit of hiring.

The author claims:

• The current gig economy is broken due to friction and low-trust dynamics.
• AI agents cannot do real-world work, but there are people who can — creating a mismatch.
• The AI era flips this dynamic: agents will create more labor, not just consume it.
• shoku aims to build a new marketplace that unleashes what agents can do, rather than shaving costs.

Inference: This is a positioning statement about reimagining the gig economy through an AI-first lens. It does not indicate any actual product adoption or market traction.

The description states:

• The system is designed for any accountable party — person, company, agent, or machine — that wants to hire against a contract.
• Customers are described as those who want to turn an outcome into a job post, without the friction of traditional platforms.

Inference: The ICP appears to be AI agents or systems looking to delegate real-world tasks. However, no specific customer segments or personas are named.

The description does not contain any information about:

• Revenue streams
• Pricing models
• Monetization strategy
• Customer acquisition costs
• Unit economics

Not evidenced

The description indicates:

• Built with Codex (OpenAI’s code generation tool)
• Uses LLMs for planning, deterministic code for execution
• Missions are structured as bounded outcomes
• Settlement is mocked; real escrow and payout are planned
• External tools resolve instantly today but need MCP adapters
• Planning runs on a reasoning model that takes seconds — future improvements include faster or streaming plans

Inference: The technical architecture suggests a hybrid approach combining LLMs for planning and typed commands for execution. However, the system is described as demo-grade with many features still under development.

The description states:

• The protocol is real, but edge features are demo-grade
• No revenue, customers or adoption data are provided
• The project was submitted to a hackathon (OpenAI 2026)

Not evidenced

The description does not mention:

• Direct competitors
• Market size or competitive landscape
• Existing solutions in the AI-agent/human-labor space

Not evidenced

• Demo-grade prototype: The system is described as largely demo-grade, with many features still under development.
• No traction evidence: No customers, revenue, or adoption data are provided.
• Unproven market demand: The positioning implies a new kind of marketplace, but no validation of that need exists in the description.
• High technical ambition without execution proof: The vision is ambitious, but there’s no indication of how the system will scale or be implemented beyond a prototype.

1. What specific real-world tasks are you planning to support with this protocol?
2. How do you plan to onboard and verify human workers at scale?
3. What is your roadmap for moving from demo-grade to production-ready?
4. Have you identified any early adopters or use cases beyond the hackathon?
5. What are the key assumptions behind the mission-based hiring model, and how might they be validated?