dengyier

Building OpenWorkProof — verifiable AI agent execution protocol. Ed25519-signed work contracts, causal evidence chains, offline replay. Apache-2.0.

China Joined on Aug 7, 2026 hbdeng113@gmail.com https://github.com/dengyier/OpenWorkProof

Education

HKU

Work

Solo Founder, OpenWorkProof

dengyier

Aug 25

The Biggest Barrier to Enterprise AI Agents Is Not Intelligence. It Is Delegation.

#ai #agents #opensource #security

8 min read

Want to connect with dengyier?

Create an account to connect with dengyier. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

dengyier

Aug 24

Human Control Cannot Be a Checkbox in an Autonomous Agent System

#ai #agents #security #opensource

7 min read

dengyier

Aug 24

A Signed AI Agent Receipt Can Still Be Wrong

#ai #agents #security #opensource

6 min read

dengyier

Aug 24

The More Autonomous AI Agents Become, the More Humans Need the Final Say

#ai #agents #security #opensource

7 min read

dengyier

Aug 24

Agent Autonomy Has a Missing Layer: Verifiable Human Authority

#ai #agents #opensource #mcp

6 min read

dengyier

Aug 20

Two Verifiers, One Verdict: Closing the 'Self-Reported Exit Code' Hole

#security #python #opensource #aiagents

4 min read

dengyier

Aug 18

I Built a Self-Evolving Knowledge Base — Here's the Architecture

#ai #obsidian #automation #productivity

6 min read

dengyier

Aug 17

I Asked 'Do You Trust an Agent That Says the Tests Passed?' — 68 Comments Later, Here's What I Learned

#ai #security #opensource #python

3 min read

dengyier

Aug 16

I Tampered With a 'VERIFIED' AI-Agent Delivery. Can You Find It in 5 Minutes?

#discuss #ai #agents #productivity

2 min read

dengyier

Aug 15

Verifying the Verifier: How an Independent Audit Hardened OpenWorkProof v0.5

#discuss #ai #agents #productivity

6 min read

dengyier

Aug 12

OpenWorkProof Protocol Specification

#agents #ai #opensource #softwareengineering

10 min read

A stray newline hid an unrunnable assert

dengyier

Aug 10

When Your AI Agent Passes 2,283 Tests — And Still Fails in Production

#discuss #ai #agents #programming

4 min read

dengyier

Aug 8

When AI Agents Ship Code: A Protocol for Verifiable Execution

#python #ai #opensource #agents

5 min read

Readers debate crypto versus simple audit trails

dengyier

Aug 7

When an AI agent says 'I ran the tests and they passed' — do you trust it?

#discuss #ai #agents #productivity

2 min read

dengyier

Aug 7

Agents can generate results — but on what authority do we accept delivery?

#ai #agents #productivity #opensource

3 min read

dengyier

Aug 7

I built a protocol so AI agents can't just say 'trust me'

#python #ai #opensource #agents

3 min read

DEV Community

dengyier

Badges

2 Week Community Wellness Streak

1 Week Community Wellness Streak

Writing Debut

Skills/Languages

Currently learning

Currently hacking on

Available for

The Biggest Barrier to Enterprise AI Agents Is Not Intelligence. It Is Delegation.

Want to connect with dengyier?

Human Control Cannot Be a Checkbox in an Autonomous Agent System

A Signed AI Agent Receipt Can Still Be Wrong

The More Autonomous AI Agents Become, the More Humans Need the Final Say

Agent Autonomy Has a Missing Layer: Verifiable Human Authority

Two Verifiers, One Verdict: Closing the 'Self-Reported Exit Code' Hole

I Built a Self-Evolving Knowledge Base — Here's the Architecture

I Asked 'Do You Trust an Agent That Says the Tests Passed?' — 68 Comments Later, Here's What I Learned

I Tampered With a 'VERIFIED' AI-Agent Delivery. Can You Find It in 5 Minutes?

Verifying the Verifier: How an Independent Audit Hardened OpenWorkProof v0.5

OpenWorkProof Protocol Specification

When Your AI Agent Passes 2,283 Tests — And Still Fails in Production

When AI Agents Ship Code: A Protocol for Verifiable Execution

When an AI agent says 'I ran the tests and they passed' — do you trust it?

Agents can generate results — but on what authority do we accept delivery?

I built a protocol so AI agents can't just say 'trust me'