Skip to content
dmae97Public

About

Open-source multi-agent coding CLI for the terminal (TypeScript, MIT). Switch models mid-session; finish goals only when your test passes.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

153 stars

Watchers

0 watching

Forks

Latest commit

 

History

19,271 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

OMK

Open Multi-Agent Kit
Make “done” pass a check.

An open-source multi-agent coding CLI for your terminal: switch models in one session and finish a goal only when your test passes.

npm version npm downloads per month
latest GitHub release CI status GitHub stars
MIT license supported Node.js version Join the official OMK Discord

60-second start · Quick start · Default or opt-in? · Context & memory · Evidence and limits
Choosing OMK · Documentation · Discord

Built with our community

Thank you to everyone who has opened a pull request directly in OMK.
Your time, ideas, and care help this project grow.

@2409324124
2409324124
@Abhinav3499
Abhinav3499
@AndreIsraelAO
AndreIsraelAO
@anhnt256
anhnt256
@AnushreeKasturi
AnushreeKasturi
@dannyward630
dannyward630
@HangMine
HangMine
@imsurya26
imsurya26
@laozhej
laozhej
@lj1166
lj1166
@Muhtasim-Munif-Fahim
Muhtasim-Munif-Fahim
@nova-loop
nova-loop
@pemalama12
pemalama12
@Shellghostt
Shellghostt
@shishir2305
shishir2305
@shoaibriazdev
shoaibriazdev
@Sudhakar7J
Sudhakar7J
@Unluckyathecking
Unluckyathecking

Open, merged, and closed pull requests · Explore OMK pull requests

OMK writes and runs code. AdaptOrch checks what it wrote. correctness_claim: false — it ran, and this is what happened.

OMK writes and runs the code. AdaptOrch checks what it wrote.
When an agent's “done” is not enough for your review, AdaptOrch keeps a record of what actually ran and what happened, and every report says correctness_claim=false.
A separate, optional product: OMK never calls it by default. How it fits | Plans | Claim boundary


CONTROL CHOICE EVIDENCE
Keep the work in your terminal and repository. Switch providers and models in one session. Define a check before a goal can complete.

OMK is a lightweight, MIT-licensed AI coding agent for your terminal, written in TypeScript. Switch models in one session, add subagents when you need them, and mark a goal done only when your own check passes. It is a standalone CLI, so it can serve as an open-source alternative to a single-provider agent such as Claude Code.

OMK 2.0 migration status: prerelease. The 2.0.0 milestone is published, but the independent Rust core is not ready to select. The default agent still runs on Node. Native omk package loading is the shipped breaking change; complete Pi removal and Rust/platform acceptance remain open. See why 2.0.0.

60-second start

You need Node.js 22.19 or newer and a git repository that has a test command.

npm install -g open-multi-agent-kit --ignore-scripts
cd your-project
omk

On the first run OMK opens /login (subscription, API key, or local model). Then, inside OMK, give it a goal and name the check:

/goal Fix the failing test without changing its assertions.
/goal verify npm test

OMK asks you to approve the command, runs it, and runs it again after each settled turn. The goal completes only when it exits 0. Swap npm test for your project's own check.

No global install? Run npx --ignore-scripts open-multi-agent-kit. The standalone binary and the setup details are in Quick start.

If omk helps you, a ⭐ on GitHub helps other people find it.

Why OMK

An AI saying “done” is not an acceptance check. With /goal verify, you choose the command that checks your goal. OMK runs it after each settled turn and completes the goal only when it passes on the current workspace. A later edit makes that evidence stale. The result is only as strong as the check you choose.

/goal Fix the failing test without changing its assertions.
/goal verify node --test check.test.mjs

Use a test that exists in your project, or follow the small, reproducible goal demo. This is an explicit workflow; ordinary prompts do not enable the gate.

Choose a model, work on your repository, then switch models with /model when another one suits the next step. The conversation stays in the same session. You can also stop and return later with /resume or omk -c.

OMK is a standalone CLI, not a plugin for Claude Code or OpenCode. It supports subscription providers, API keys, and local models. Start with one agent that reads files, edits code, and runs commands. Add subagents or explicit verification workflows when you need them; neither is required for your first task.

Why 2.0.0 and what changed

2.0.0 marks the move to OMK-native packages by default, rather than silently loading legacy Pi manifests and substituting Pi SDK imports. That breaks some existing extension installations, so the owner declared a milestone version. It does not mean the Rust rewrite is complete. The migration remains prerelease; GitHub currently lists the published v2.0.0 milestone as a non-prerelease, and npm has 2.0.0. Those publication labels are not Rust-core acceptance evidence.

Area What shipped in the 2.0.0 milestone What remains
Native packages The default manifest dialect is omk; Pi-only manifests and imports that relied on implicit aliases no longer load by default Migrate extension manifests and imports, then check them with omk package doctor <source>. Native mode is not proof that every Pi dependency is gone
Extension SDK Session-scoped models and registerEntryRenderer() for display-only persisted entries; optional omk-speak-like-you-eat companions Existing extensions still need migration and compatibility checks
Suggestions and skills Next-task suggestions with Tab insertion (OMK_NEXT_SUGGESTIONS, flag, default off); clause-aware automatic skill selection These do not enable multi-agent execution or verify an answer
Rust core K0 contract types and generated TypeScript declarations Provider, loop, tool and extension-host migration, the K5 headless candidate, and packaged-artifact acceptance on Linux, macOS and Windows

For native extensions, rename the pi manifest key to omk and replace legacy SDK imports with the matching open-multi-agent-kit, omk-ai, omk-tui or omk-agent-core API. Do not enable omk+pi just to hide migration failures. The explicit compatibility adapter remains available for installations that choose it; it is not the native migration path or the default.

The standalone download bundles the current JavaScript runtime. Needing no separate Node installation does not make it a Rust-owned agent core. The opt-in Linux Rust sandbox supervisor is also a separate, bounded capability, not the full core. The Rust terminal UI and remaining subsystem ownership follow the headless candidate in the existing staged plan. Upstream copyright and license notices remain required throughout the port.

See the migration guide, Rust core plan and published milestone notes for the scope and gates. This status note was checked on 2026-11-14; it does not retag or republish 2.0.0.

At a glance

  • Done means a passing check. /goal verify <command> completes a goal only when your check passes on the current workspace.
  • One session, many models. Switch with /model mid-conversation; resume later with /resume or omk -c.
  • Sandboxed by default. Built-in bash runs under sandbox-exec or bwrap, without network, and fails closed.
  • Opt-in power. Subagents, MCP servers, skills and extensions load only when you configure them.
  • Honest evidence. Claims below link to the docs that back them, and the limits are stated next to them.
flowchart TD
  P["Prompt or goal"] --> A["Agent loop"]
  A --> S["Settled turn"]
  S --> R["Response"]
  S -->|"goal verification configured"| V["Run acceptance check"]
  V -->|"pass on current workspace"| D["Goal complete"]
  V -->|"fail"| A
Loading

The loop uses your selected model and available tools. Subagents, MCP, skills, and extensions depend on your configuration.

Quick start

Install from npm (Node.js 22.19 or newer), then start in your repository:

npm install -g open-multi-agent-kit --ignore-scripts
cd your-project
omk

Or run npx --ignore-scripts open-multi-agent-kit without a global install.

Or install the standalone binary, which needs no Node.js (macOS, Linux):

curl -fsSL https://github.com/dmae97/omk/releases/latest/download/install.sh | sh

The script checks the download against the release's SHA256SUMS and installs to ~/.omk/bin; add that folder to your PATH as it tells you. Windows builds are attached to each release.

  1. On a first run OMK opens /login for you: sign in with a subscription or an API key. An existing Claude Code or Codex CLI login can be reused with omk provider adopt.
  2. Run /model to choose another available model at any time.
  3. Try a read-only first task:
Summarize this repository and identify the commands used to check it.
Read the project configuration to support your answer. Do not edit files.

After the reply, use /model to choose another configured model and ask it to review the answer. You stay in the same session. This is manual model switching, not parallel agents or an independent correctness check.

Ready to see a test-backed goal? Try the goal demo, then share your result or a reproducible failure in a GitHub issue. If this workflow is useful, star OMK to help other developers find it.

For a bug fix, name the failing behavior and ask for a regression test, the smallest fix, and the check commands with their exit codes. Review the diff and those results yourself; a request to run tests does not enable a verification gate.

Built-in local bash requires sandbox-exec on macOS or bwrap plus unprivileged user namespaces on Linux (sudo apt install bubblewrap on Debian and Ubuntu). It blocks network access and fails closed if the backend is missing. omk doctor checks the runtime, credentials, the model a first session would pick, and the sandbox, and prints the fix for this machine. See Bash sandbox setup, the safety boundary and the full quickstart.

What runs by default

A fresh install starts one agent/tool loop after provider setup. It does not turn each prompt into a multi-agent workflow or automatically certify its answer.

Capability Fresh-install behavior Where to start
File editing, shell commands, saved sessions Built in; tools run when called by the agent Usage, sessions
Tool-call scheduling dag-v2 schedules resource conflicts within the agent loop; it does not launch a team Runtime algorithms
Subagents Optional extension; load it and supply agent definitions Subagent setup and examples
MCP servers, extra skills and extensions Require configured servers or installed resources MCP, skills, extensions
Durable goals Built in; /goal <objective> continues after each settled turn, up to 8 rounds. With /goal verify <command> the goal completes only when that check passes on the current workspace Acceptance checks
Protocol verification and advisory judging Explicit API/workflow opt-in; not a gate on ordinary prompts Run protocol
Verified runs (omk run) Opt-in CLI and SDK on Linux with bwrap; runs an approved command in an isolated copy, checks the result, and supports resume, cancel and cleanup Verified run
Context budgeting Off by default Settings
AdaptOrch integration Optional and separate; no service calls by default OMK + AdaptOrch

The internal lane launcher and automatic command-sharding primitives are not connected to the default CLI path. Installing their packages is not the same as enabling an orchestration workflow.

Context, cache and memory

OMK manages several distinct resources. A prompt-cache boundary, a local cache, and the process's RAM are not the same thing; provider cache hits and costs depend on the provider and the request.

Layer What OMK does Default and scope
Prompt prefix Preserves the cache boundary when an extension changes only the dynamic tail and the prefix matches byte for byte. On by default; this does not guarantee a provider cache hit. OMK_PREFIX_BOUNDARY
Context Budget V2 Reserves response and safety space, keeps required items, and selects representations of optional context that fit. Off by default; enable globally or for a process. Context selection
Local context cache Reuses eligible representation and negative-result entries; cached plans stay in session memory. Only when Context Budget V2 is enabled. Representations persist per workspace by default; a memory-only option is available. Settings
Session history Compaction can summarize earlier context; saved sessions support resume. Compaction and its settings are separate from persistent knowledge and provider-side caches. Sessions
Host resources The resource governor can observe host capacity and admission decisions. observe is the default; adaptive throttling requires configuration. Resource governor

For one process, opt into context budgeting and keep its cache in session memory:

OMK_CONTEXT_GOVERNOR=1 OMK_CONTEXT_GOVERNOR_CACHE=memory omk

Memory-only changes where entries are kept; it is not a guarantee of lower RAM use. Cache effectiveness, memory use, latency and cost depend on the workload. Measure them before claiming an improvement; turn metrics show what can be recorded.

Evidence and limits

One dated comparison is available below: OMK and mini-SWE-agent on Terminal-Bench 2.1 with the same base model. The observed success-rate difference is not statistically significant, and the time budgets differed. It does not establish overall harness superiority, multi-agent gains, or how much verification reduces false completion.

OMK targets state-of-the-art quality as a CLI coding-agent harness. SOTA is not verified.

The evidence you can inspect today covers specific failure modes:

Behavior covered Regression evidence Scope
Missing test observations produce inconclusive; a required failing test produces fail Protocol tests Explicit protocol evaluation, without a waiver
Changed artifacts, wrong command bindings, or missing ledger evidence block acceptance Evidence binding tests Strict evidence gate and selected workspace scope
A relevant workspace mutation after verification makes the receipt stale Freshness tests Configured receipt and mutation tracking
A durable goal with an approved acceptance check completes only on a passing receipt, and an edit after the check makes that receipt stale Goal acceptance tests, live loop tests Approval held by the running OMK process; git work tree; static command lines
An effect that may still be live keeps its resource claims through expiry, cancellation and authority restart, until a supervisor confirms termination Coordination broker tests In-process broker with canonical claim keys; no OS fencing
A publication is refused unless its read versions, parent revision and receipt binding all still match Publication tests Single accepted snapshot pointer; no multi-file filesystem atomicity
Cancellation after dispatch is never reported as cancelled-before-dispatch; the outcome stays unknown until settled Operation lifecycle tests Pure state machine; does not itself stop a remote effect
Zero trials is reported as absent evidence rather than zero risk, and a point estimate is not an error rate Risk bound tests Binomial model under a fixed policy and adequately independent samples

These tests exercise the gates, not the rate at which they catch real bugs. An ordinary prompt finishes when its tool loop and queued work settle; prompt_settled is not a correctness verdict.

A useful comparison must hold the model, provider configuration, tasks, budget, and tool permissions constant, and label default versus opt-in workflows. Report task success, cost, latency, and false completion (reported complete but failing the declared checks), with its denominator and per-task outcomes. The measurement protocol defines the reproducibility and privacy requirements. omk stats shows local turn costs and tool failures; it does not score task correctness.

If you evaluate OMK, share a sanitized report and reproduction steps in a GitHub issue. Include failed and interrupted runs, not just successful examples.

Terminal-Bench 2.1 · R8

Self-measured result OMK evaluation build mini-SWE-agent 2.4.6
Observed success 75.8% (200/264) 71.6% (189/264)
Recorded cost per trial $0.581 $0.759
Median wall time per trial 482 s 376 s

75.8% observed success · approximately 23% lower recorded model cost per graded trial. OMK took longer: 482 s median, compared with 376 s for mini-SWE-agent. Both used xAI grok-4.7 at reasoning effort xhigh in a 2026-10-05–06 KST run of 89 tasks, with 3 trials per task per harness.

The +4.2 percentage-point success difference is not statistically significant (95% task-level paired-bootstrap CI: −1.9 to +10.2 pp). OMK alone had 150 s extra per trial for snapshots, so the time budgets were unequal. This result covers the 1.3.0 evaluation build 78cc483, including unmerged changes, rather than the published npm package. It is not an official leaderboard result.

Benchmark data and evaluation limits
Harness Success rate (88 tasks × 3 trials) Cost per trial (all 267 trials) Median wall time per trial (all 267 trials)
OMK 75.8% (200/264) $0.581 482 s
mini-swe-agent 2.4.6 71.6% (189/264) $0.759 376 s
  • The difference is +4.2 percentage points with a 95% confidence interval of [−1.9, +10.2] (task-level paired bootstrap). The interval includes zero, so this is not a statistically significant improvement.

  • OMK cost about 23% less per trial but was slower: 482 s median per trial against 376 s.

  • Audit corrections: trials that tried to read Terminal-Bench sources or answers score 0, which turned 4 mini-swe-agent passes and 1 OMK pass into failures. prove-plus-comm is excluded from both sides because our OMK adapter failed to start in that task's working directory. Two transient provider errors were retried; no trial was excluded.

  • OMK failed all three pytorch-model-recovery trials because the CLI rejected an instruction starting with a dash and a space (fixed in #84). Without that task the difference is +5.4 pp [0.0, +10.7].

  • Not a leaderboard run: 3 trials instead of the official 5, our own network allowlist and apt cache proxy, and OMK alone had 150 s extra for a workspace snapshot used in rescoring.

  • Costs count only the graded attempt, excluding prior failed retry attempts and 42 auxiliary model calls made by code inside one mini-SWE-agent task. They are not total experiment spend.

Method, per-difficulty results, audit and reproduction notes: Terminal-Bench 2.1 report.

Editable figure · Original report review

OMK//CONTROL

The terminal UI shows the selected model, tools, and session status. Additional signals depend on the integrations you configure.

In the terminal What it shows
Model Selected provider and model
Tools Available tools and their activity
Session Status and the current conversation

The header reads omk v<package.version> · OMK//CONTROL; the installed package version is the source of truth.

Control loop

How scope, routing, verification, and replay fit together

OMK's provider-neutral coding-agent CLI also exposes a multi-agent control plane for explicitly configured workflows. The stages describe that design, not what every prompt automatically runs.

01 SCOPE 02 ROUTE 03 VERIFY 04 REPLAY
Define the task and boundaries. Choose models and configured tools. Run the declared checks. Retain state and evidence for recovery.

The v0.98.3 SDK rejects incomplete first-party judge responses and exposes deterministic ties; it is not an automatic TUI judge.

  1. Scope the goal, paths, resources, and acceptance predicates. A selected orchestration workflow may also supply a DAG; an ordinary prompt remains one agent/tool loop.
  2. Route work to models, agent skills, MCP tools, and extensions without changing the evidence contract.
  3. Verify declared checks in explicit evidence workflows. Required failing checks block those workflows; advisory judging cannot replace them.
  4. Preserve receipts and replay state for bounded session recovery; continue durable goals from explicit reducer state.

Verification boundary

AgentSession built-in local bash uses OS sandbox enforcement by default: sandbox-exec on macOS and bwrap plus unprivileged user namespaces on Linux. Local shell spawns restrict writes to the workspace and OS temporary directory, disable network access, and fail closed with sandbox.backend_missing when an enforcement backend is unavailable.

This is not read-confidentiality or whole-process containment. Other file tools, extension and custom-tool code, injected or remote BashOperations, and the OMK process keep the permissions of the process running them. Use containerization when the boundary must cover more than built-in local bash. Explicit evidence workflows cannot treat missing required evidence as a verified result.

Providers

Provider integrations and published packages

OMK keeps routing separate from control and evidence. Codex, Claude Code, OpenCode Zen/Go, Kimi, GLM/ZAI, native xAI/Grok, NVIDIA NIM, and local providers can participate through omk-ai while the run contract stays stable.

Native xai keeps subscription OAuth and XAI_API_KEY billing separate. See provider setup, provider resilience, and Grok integration.

Published packages

Package Purpose
open-multi-agent-kit Interactive coding-agent CLI and control plane
omk-agent-core Agent runtime, tool execution, and DAG scheduling
omk-ai Unified multi-provider LLM API
omk-protocol Versioned run contracts and semantic reducers
omk-adaptorch-wpl Work Packet Loop runtime
omk-book-to-skill Optional document-to-skill compiler
omk-tui Differential-rendered terminal UI library
npm install omk-agent-core
npm install omk-ai
npm install omk-protocol
omk install npm:omk-book-to-skill@0.98.3
npm install omk-tui

Repository understanding

Optional indexes, retrieval settings, and trust limits

v0.97.0 shipped the OpenWiki policy and workflow, but no versioned corpus or integrity checker. The following integrity/output guards shipped in v0.98.0; the generated corpus remains optional and is not bundled:

  • openwiki/ — absent. The previous untracked corpus was removed after the hardened gate proved it carried fabricated evidence: 8 frontmatter symbols that no declared source path defines (AgentLoop, getModel, DeepWall, loadExtensions, createExtensionRuntime, main), 45 references to @omk/* package names this repository does not publish, and a restatement of this README's Scope -> Route -> Verify -> Replay loop as a strict engine state machine, which is not what the source implements. CI regenerates the corpus; nothing is lost.
  • scripts/check-openwiki.mjs — shipped integrity checker. An interrupted corpus now fails unless openwiki/.manual-review.json binds a review to the exact corpus digest, and every frontmatter symbol must bind to one of that page's own source_paths as a whole identifier.
  • scripts/check-openwiki-output.mjs — output gate. The scheduled workflow may write, upload, and open a PR for openwiki/ and nothing else, so a model reading this repository cannot reach AGENTS.md, CLAUDE.md, or the workflow that runs it. The gate runs once before the artifact leaves the read-only generating job and again before the PR, because the publishing job holds write permissions the first one does not.
  • .understand-anything/ — optional local structural graph used by Pi Lens; it is not published or injected into prompts by default. To reach a session, attach it through OMK's MCP client like any other server; there is no second, bespoke path for it.

Source and tests remain authoritative. Shipped guards do not turn a generated index into authority: treat corpus pages as local advisory data and recheck source.

Retrieval

A corpus no session can read is documentation of a plan, not a feature, so the pages are now candidates for prompt budgeting. Enable contextBudget.openwiki alongside contextBudget.enabled (settings) and each page becomes a low-priority evidence item ranked against the turn's query. Pages compete for leftover budget and can never displace instructions or skills; most turns carry titles and declared symbols alone, and a page's text arrives only when the query earns it.

Admission mirrors scripts/check-openwiki.mjs rather than restating it. A complete corpus at the current HEAD offers page text; one whose HEAD has moved offers titles only and is marked stale; an interrupted corpus is refused unless a review binds to its exact digest. The default is off, and with the setting off the prompt is byte-identical to one built without a corpus.

Prior art

Research references, not OMK benchmark results

The design decisions behind OMK's context, routing, memory, and orchestration layers are grounded in published work rather than invented in isolation. Each row below was retrieved and read directly; claims are at abstract level, which is the evidence grade this table asserts and no more.

Paper Mechanism it establishes OMK implementation or design reference
arXiv:2608.22752 — The Compaction Cliff in Long-Running AI Agent Memory Uniform summarization erodes rules and episodic logs at the same rate; measured safety-rule retention falls to 53% after one compaction and 10% after five. Type-tagged deterministic operators fix it. Type-aware compaction triage: rule-typed items survive N rounds byte-identical
arXiv:2608.23023 — Most of the LLM Routing Gap Is Task Type Most routing gain is reachable with a fixed task-type table; run-to-run flips must not be credited as wins. Frozen task-class table plus the 2-run stability rule in the promotion gate
arXiv:2506.16655 — Arch-Router: Aligning LLM Routing with Human Preferences Indirection: a classifier emits a label, a policy table maps label to decision, so models change without retraining. classifyTaskV4 plus TASK_CLASS_THINKING_LEVELS
arXiv:2605.09894 — Deterministic vs. LLM-Controlled Orchestration Holding model, prompts, and tools constant and varying only execution control, deterministic orchestration matched accuracy, improved worst-case robustness, and cut tokens up to 3.5x. Deterministic scheduler and planned lanes; execution control is never delegated to the model
arXiv:2608.15565 — Admission Without Answers Label-free admission on execution success alone admits substantial contamination; an accept/abstain/escalate decision is required. Verified-memory admission design (spec 019), abstain is not stored
arXiv:2608.23471 — InjecMEM: Memory Injection Attack on LLM Agent Memory Systems Single-interaction memory injection is a reproduced attack frame against agent memory. Retrieved memory is injected only as provenance-tagged data, never fused into instruction position

Entries include implemented mechanisms and design proposals; check the runtime status guide for availability. The wider survey, including approaches not adopted, is working material that is not published with the repository.

Documentation

Development

npm ci --ignore-scripts
npm run build
npm run check
npm test
npm run release:local

Direct dependencies are pinned, CI installs with --ignore-scripts, and the published CLI includes a generated npm-shrinkwrap.json. Read CONTRIBUTING.md and the development guide before sending a change.

FAQ

What is OMK?

OMK (Open Multi-Agent Kit) is an open-source coding agent CLI for the terminal. It reads files, edits code, and runs commands in your repository. You can switch models inside one session, add subagents when a task needs them, and use /goal verify so a goal only counts as done when your chosen check passes. It is MIT-licensed and written in TypeScript.

How do I install OMK?

With Node.js 22.19 or newer, run npm install -g open-multi-agent-kit --ignore-scripts, then omk in your project. Without Node.js, use the standalone installer on macOS or Linux: curl -fsSL https://github.com/dmae97/omk/releases/latest/download/install.sh | sh. See Quick start.

Why use OMK instead of Claude Code?

Use it for provider choice within one CLI session, or to build workflows against its public runtime and evidence APIs. For a single-provider workflow, your current agent may be sufficient. Try the read-only task above before moving existing work.

How is this different from OpenCode with plugins?

OMK is a separate runtime with its own CLI, sessions, tool scheduler, and SDK. One reason to choose it is to build your own acceptance workflow: define required test observations in the run protocol, then have your automation reject fail or inconclusive results. Receipt integrity and freshness still need their own configured checks.

For adding a tool or prompt to an existing OpenCode setup, a plugin may be the smaller change. OMK's protocol is opt-in, not proof of better performance.

Does multi-agent mean better results or automatic verification?

Not automatically. Subagents require setup, and verification must be part of the chosen workflow. Its result covers the declared checks, not all behavior. See what runs by default and evidence and limits.

Recent releases

Release notes and historical corrections

Historical correction: the immutable v0.97.0 notes below announced a versioned OpenWiki corpus, but that release still ignored /openwiki/ and did not contain the corpus or checker. See the current repository-understanding section above for the shipped guards and optional-corpus boundary.

Release v2.0.1

Fixed

  • omk init --global and a bare omk launch now report why bundled pstack agent installation failed, including the affected path, instead of (unknown). Existing files are still never replaced (#451).
  • An extension load failure for a module under another installation's node_modules tree no longer suggests omk install --allow-scripts, because install scripts cannot fix a stale path. A missing file in a package's own build/, dist/ or lib/ output and missing .node bindings still get the hint (#451).
  • Skill selection rejects fuzzy matches between words with different initial characters unless one contains the other, and no longer treats path-only fragments as independent capability clauses. Filename tokens still count toward whole-request relevance (#453).

Changed

  • Next-task suggestions (OMK_NEXT_SUGGESTIONS, flag, default off) use the v3 format: a title and three labelled lines for context, scope and completion criteria. Skill selection for an accepted suggestion uses its title and scope rather than the host's context and check wording (#454).

Release notes live in RELEASE_NOTES_v2.0.1.md.

Release v2.0.0

Added

  • omk-speak-like-you-eat adapts SLYE 1.6.1 as an optional OMK package with manual and automatic plain-language companions, ELI5 and Light tabs, scoped model selection, trusted project settings, custom prompts, cancellation, and retry behavior. Extension contexts now expose session-scoped models, and registerEntryRenderer() renders persisted custom entries without adding them to LLM context. See Speak Like You Eat and Extensions.
  • Next-task suggestions (OMK_NEXT_SUGGESTIONS, flag, default off) now show a complete prompt in the empty editor: a title, then input, verification, evidence and result sections rendered by the host from the session transcript. A local suggestion appears before any provider call and stays when the model reply is empty, malformed, timed out or failed; Tab inserts exactly the displayed text and refuses a suggestion whose session state changed. Multi-line editor placeholders render below the cursor row. See environment variables.

Changed

  • Skill selection for automatic loads and provider-harness grants scores each clause of a multi-capability request and grants one skill per clause before a second for the same clause, still at most three. Single-clause requests rank as before. See skills.

Breaking Changes

  • The default package manifest dialect is now native omk. Extensions that import legacy Pi packages without installing them, and packages with only a pi manifest, no longer load by default. Opt into compatibility with --package-manifest-dialect omk+pi, or packageManifestDialect: "omk+pi" in SDK constructors. The load error now names this option. omk package doctor also defaults to the native target. See the OMK 2 migration notes.

Fixed

  • A receipt-reading critic, visual-QA or document lane now depends only on the producers whose receipt its read scope covers; a lane reading the whole run directory still waits for every producer. Writer serialization and the failure scope are unchanged.

Release notes live in RELEASE_NOTES_v2.0.0.md.

Release v1.5.14

Fixed

  • Turns on Anthropic models gated to Claude Code 2.1.290 or newer no longer fail with HTTP 400 claude_code_version_too_old when you are logged in with an Anthropic OAuth subscription. The custom-provider-anthropic example extension sends the same version in its user agent (#393).

Release notes live in RELEASE_NOTES_v1.5.14.md.

Contributing

Issues and pull requests are welcome. Read CONTRIBUTING.md first, run npm run check and npm test before opening a pull request, and report security issues as described in SECURITY.md.

Star OMK

If a check-backed “done” is useful to you, a star on GitHub helps other developers find OMK.

Star history chart for dmae97/omk

Acknowledgments

OMK builds on pi — Mario Zechner's MIT-licensed coding-agent harness — and began from the oh-my-pi fork. The vendored tree was removed in this release line; OMK 0.9x is OMK-native (see specs/constitution.md), and the design debt to both projects stands. Thank you.

OMK + AdaptOrch

AdaptOrch is a separate, optional product and is not called by default. See OMK + AdaptOrch for the optional integration and TRADEMARKS.md for the name and marks.

License

MIT

About

Open-source multi-agent coding CLI for the terminal (TypeScript, MIT). Switch models mid-session; finish goals only when your test passes.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

153 stars

Watchers

0 watching

Forks

Releases

Sponsor this project

Packages

Contributors

Languages