Skip to content
View aDragon0707's full-sized avatar
15 followers · 12 following

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Block or report aDragon0707

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
aDragon0707/README.md

Evan Ma

每天进步一点点，探索人与 AI 共同进化。
Improving a little every day, exploring human-AI co-evolution.

I build auditable workflows for AI agents: task packets, work receipts, handoffs, and human review boundaries.

My current focus is SACP, a lightweight protocol for making long-running AI agent work traceable, reviewable, and improvable.

Start Here

  • SACP
    State-Aware Collaboration Protocol for auditable AI agent work.

  • Solo AI Company OS
    A Markdown-first operating system for human-AI collaboration: decisions, roles, worklogs, handoffs, and review records.

  • token-prompt-compiler
    Turns messy human intent into bounded, token-efficient task packets for LLM agents.

  • Agent Flight Recorder
    Turns agent runs into snapshots, evidence trails, improvement cards, and field notes.

  • read-research-papers
    A Codex skill implementing S. Keshav's three-pass method for progressively deeper paper reading.

  • Portfolio
    My bilingual personal portfolio for AI systems, learning, and selected work.

What I Care About

  • AI agent accountability
  • human-in-the-loop workflows
  • task packets and work receipts
  • prompt and context engineering
  • evaluation evidence and dirty-run cases
  • AI-assisted learning systems
  • solo AI company workflows

Direction

I use GitHub as a public lab for turning AI workflow experiments into clearer protocols, tools, demos, and case studies.

The long-term question I care about:

How can humans and AI systems work together in a way that is traceable,
reviewable, improvable, and aligned with human judgment?

Links

Pinned Loading

  1. audit-evolution-agent-flight-recorder audit-evolution-agent-flight-recorder Public

    Agent Flight Recorder skill for snapshots, evidence trails, evolution cards, and field notes.

    HTML 3

  2. claude-code-html-skill claude-code-html-skill Public

    Codex/Claude Code skill for official-style HTML artifacts with Markdown and Obsidian round-trips

    HTML 4 1

  3. sacp sacp Public

    State-Aware Collaboration Protocol: auditable work receipts for AI agents.

    Python 14

  4. Solo-AI-Company-OS Solo-AI-Company-OS Public

    Markdown-first operating system for human-AI collaboration: decisions, roles, worklogs, handoffs, and reviews.

    Python 4 1

  5. token-prompt-compiler token-prompt-compiler Public

    Compile human language into token-efficient machine-readable task packets for LLM agents

    JavaScript 2