meta-kim

Language: English | 简体中文 | 日本語 | 한국어

Start here

Tell Meta_Kim the outcome you want. It helps choose an evidence-backed approach, find the right capabilities, and coordinate work through verification.

Meta_Kim is a general agent governance framework that runs through hosts such as Claude Code and Codex. Its core is intent, decision-making and orchestration across research, operations, learning, writing, software and other tasks supported by discovered capabilities. Industry roles are optional capabilities. The framework compares feasible routes, coordinates owners and dependencies, verifies outcomes, and preserves useful lessons.

Current release: v3.7.3. Release notes: English · 简体中文 · 日本語 · 한국어.

You say: “Compare the feasible options under my budget and time constraints, recommend one with evidence, organize the work, and tell me which outcomes were actually verified.”

What you actually get

• The thing you asked for: a decision with evidence, a research report, an executable plan, learning materials, code, a review, documentation, or a release-ready change.
• A readable explanation of what changed and why.
• Test and verification evidence, with unknown or unproven claims kept visible instead of being called “done.”
• A local Live control room that leads with the current task, available results and blockers, with an execution graph, work board, evidence and replay history for closer inspection.

What it looks like while running

Live leads with the current task, available results, blockers and observed activity. Named sessions, the dependency graph and its Inspector let you inspect ownership and evidence; the four-column work board and project/session repository view remain available. Task progress follows durable events, while returned calls and producer self-checks remain distinct from accepted results.

For an explicitly bound local Codex project, conversation and turn, Live can read the selected turn and verified child-task records within bounded reads. It projects structural execution summaries, without publishing raw prompt or response text. Omitted history stays visibly partial; this does not claim equivalent observation of arbitrary Claude Code conversations.

Recent upgrades on main

What the final handoff looks like

Done      Sign-up, login, token refresh, and error states implemented
Changed   8 source files, 3 tests, 1 migration
Verified  Unit 42/42 · integration 11/11 · build passed
Open      OAuth browser acceptance was not run — credentials unavailable
Next      Connect staging credentials, then replay the blocked check

That summary is backed by the changed project files and run evidence. Meta_Kim does not turn an unavailable check into a green result.

What can it help you make?

The coding runtime still does the domain work. Meta_Kim makes the route, ownership, review, and proof explicit around it.

Try it in 3 steps

1. Install Meta_Kim:

   npx --yes github:KimYx0207/Meta_Kim meta-kim
   

2. Open your project in Claude Code or Codex and ask normally. You do not need to memorize a command:

   Review this project, fix the three highest-impact problems,
   use parallel work where it is safe, and verify the final result.
   

3. See the run. The first governed request lazily starts one local Live Hub and returns the current graph link in the conversation. The manual recovery command remains:

   meta-kim live
   

Good first requests include building a feature across frontend/backend/tests, investigating and fixing a difficult bug, reviewing and improving a pull request, turning a product idea into a PRD and first implementation, or preparing a change for release without hiding failed checks.

In one sentence: Claude Code and Codex are the hands; Meta_Kim is the team structure, quality control, and evidence trail around those hands.

Technical architecture, runtime support, and maintainer commands

Overview

Meta_Kim is not another AI coding tool. It is a governance layer for durable AI coding work.

Meta_Kim Live: proof you can see

Meta_Kim Live turns durable run records and explicitly bound local Codex observations into a local Proof-Carrying Run Graph. One user-level Hub lists explicitly registered Meta_Kim projects, then lets you choose a Session / Run and inspect its tasks, results, stages, owners, evidence, uncertainty, and replay. Codex observation is read-only, limited to the selected turn and verified child-task records, and projects summaries rather than raw conversation text; omitted history stays partial. It never scans the whole disk, replaces the agent, creates a second scheduler, or pretends an unverified node is done.

The Hub opens in Chinese by default; use the EN switch in the top-right corner for English. The preference is remembered locally. A marker-backed project joins the catalog when you explicitly open Live for it or start its first governed run, while an existing explicit skip is preserved.

meta-kim live

The Hub binds only to 127.0.0.1, loads no third-party assets, and reuses one process through a PID, process-start identity, and instance-health proof; first use after an update safely replaces an older Hub version. The default surface is read-only; guarded controls are opt-in only with --enable-control and a complete injected authority loadout (durable repository, lease/fence/effect checks, control token, and capable adapter). Without that complete authority chain, the service remains plan-only and fail-closed. Selecting and reopening an observed run is available now; automatic execution resume and cross-runtime handoff remain separate capabilities that require lease/fence and side-effect safety evidence.

M3-L02–L04 sharing and control boundaries