{"slug": "oh-my-openagent", "title": "Oh-My-Openagent", "summary": "The open-source project oh-my-openagent, maintained by developer Yeongyu Kim and AI assistant Jobdori, announces support for OpenAI's Codex CLI via a new LazyCodex installer and a multi-harness refactor to support OpenCode, Codex, Pi, and others. The project claims Anthropic blocked OpenCode because of its usage, positioning itself as an orchestrator across models rather than a single provider. Testimonials praise its speed and capability, with one user reporting fixing 8,000 ESLint warnings in a day.", "body_md": "Note\n\n**OmO for Codex is here: try LazyCodex**\n\nWe loved Anthropic models enough to get blocked. Now we are backing Codex. If you are an OmO fan but the setup felt like too much, use LazyCodex. OmO for Codex has shipped:\n\n```\nnpx lazycodex-ai install\n```\n\nLearn more at [lazycodex.ai](https://lazycodex.ai).\n\nNote\n\n**Multi-Harness Agent OS Refactor in Progress**\n\nWe are restructuring the codebase to support multiple agent harnesses (OpenCode, Codex, Pi, and others). If you are interested in contributing, please read the [ROADMAP](/code-yeongyu/oh-my-openagent/blob/dev/ROADMAP.md) first. PRs related to roadmap work should use the `ROADMAP`\n\nlabel.\n\nTip\n\n**Building in Public**\n\nThe maintainer builds and maintains oh-my-openagent in real-time with Jobdori, an AI assistant running on a heavily customized fork of OpenClaw. Every feature, every fix, every issue triage — live in our Discord.\n\nNote\n\nOmO is maintained by Jobdori, the AI assistant shown above. Meet your own Jobdori — Dori.\n\nJoin the waitlist[here].\n\nTip\n\nBe with us!\n\n| Join our\n`oh-my-openagent` users. |\n|---|\n\n`oh-my-openagent`\n\nused to be posted on my X account. Since it was mistakenly suspended,\n\n[@justsisyphus](https://x.com/justsisyphus)now posts updates on my behalf.[@code-yeongyu](https://github.com/code-yeongyu)on GitHub for more projects.This is oh-my-openagent, running Team Mode. With Kimi K3 and GPT-5.6 Sol.\n\nAnthropic\n\nblocked OpenCode because of us.Yes, this is true.They want you locked in. Claude Code is a nice prison, but it's still a prison.You don't need to pay $200 for 2 hours of work. The future isn't picking one winner; it's orchestrating them all. Models get cheaper every month. Smarter every month. No single provider will dominate. We're building for that open market, not their walled gardens.\n\n\"It made me cancel my Cursor subscription. Unbelievable things are happening in the open source community.\" -\n\n[Arthur Guiot]\n\n\"If Claude Code does in 7 days what a human does in 3 months, Sisyphus does it in 1 hour. It just works until the task is done. It is a discipline agent.\"\n\n- B, Quant Researcher\n\n\"Knocked out 8000 eslint warnings with Oh My Opencode, just in a day\"\n\n-[Jacob Ferrari]\n\n\"I converted a 45k line tauri app into a SaaS web app overnight using Ohmyopencode and ralph loop. Started with interview me prompt, asked it for ratings and recommendations on the questions. It was amazing to watch it work and to wake up this morning to a mostly working website!\" -\n\n[James Hargis]\n\n\"use oh-my-opencode, you will never go back\"\n\n-[d0t3ch]\n\n\"I haven't really been able to articulate exactly what makes it so great yet, but the development experience has reached a completely different dimension.\" -\n\n[苔硯:こけすずり]\n\n\"Experimenting with open code, oh my opencode and supermemory this weekend to build some minecraft/souls-like abomination.\" \"Asking it to add crouch animations while I go take my post-lunch walk. [Video]\" -\n\n[MagiMetal]\n\n\"You guys should pull this into core and recruit him. Seriously. It's really, really, really good.\"\n\n- Henning Kilset\n\n\"Hire @yeon_gyu_kim if you can convince him, this dude has revolutionized opencode.\"\n\n-[mysticaltech]\n\n\"Oh My OpenCode Is Actually Insane\" -\n\n[YouTube - Darren Builds AI]\n\nYou're juggling Claude Code, Codex, and random OSS models. Configuring workflows. Debugging agents.\n\nWe did the work. Tested everything. Kept what actually shipped.\n\nInstall oh-my-openagent. Type `ultrawork`\n\n. Done.\n\noh-my-openagent ships in three editions of the same product: two plugins that load into a host you already run, plus one standalone edition.\n\n**Ultimate Edition (omo for OpenCode)**— full omo. 11 agents, 54+ lifecycle hooks, 5 built-in MCPs, all slash commands, Team Mode,`/goal`\n\n, ultrawork — everything. Hashline edits are opt-in (`hashline_edit: true`\n\n).**Light Edition (omo for Codex CLI)**— the portable components that fit Codex's plugin system:`rules`\n\n,`comment-checker`\n\n,`git-bash`\n\n,`lsp`\n\n,`ultrawork`\n\n,`ulw-loop`\n\n,`ulw-execute-continuation`\n\n, and`telemetry`\n\nat the core, plus`teammode`\n\nand supporting components (`bootstrap`\n\n,`lcx`\n\n, and more), plugin-scoped MCPs for`grep_app`\n\n,`context7`\n\n,`git_bash`\n\n, and`lsp`\n\n, and the shared`ast-grep`\n\nskill. It installs Codex agent TOMLs into`~/.codex/agents/`\n\n. No OpenCode`team_*`\n\ntools — Codex CLI's own spawn/collaboration surface does that work.**Senpi Edition (standalone, beta)**— the native`omo`\n\ncommand with the OMO extension built in. It installs from`omo-ai@beta`\n\nrather than loading into OpenCode or Codex.\n\nPick the edition(s) you want.\n\n| You want | Run | What lands on disk |\n|---|---|---|\nUltimate (OpenCode) |\n`bunx oh-my-openagent install` (TUI walks you through it) |\nPlugin registered in `opencode.json` + agent/model config + provider auth prompts |\nLight (Codex CLI) |\n`npx lazycodex-ai install` |\n`~/.codex/plugins/cache/sisyphuslabs/omo/` + local Codex marketplace cache + `~/.codex/config.toml` marketplace/plugin/agent blocks + optional autonomous permissions + component CLIs in `~/.local/bin` |\nBoth |\n`bunx oh-my-openagent install --platform=both` |\nBoth of the above |\nSenpi edition (beta) |\n`npm i -g omo-ai@beta` , then `omo` |\nThe `omo` command: pinned senpi release with the OMO extension built in. Beta channel only; a bare `npm i -g omo-ai` fails by design. See the\n|\n\n`lazycodex-ai`\n\ndefaults to the Codex Light installer and runs through Node/npm. `--platform`\n\non the shared `omo-agent-toolkit`\n\nCLI still defaults to `opencode`\n\n(Ultimate).\n\n- Already use OpenCode, or want the most-tested path? Choose\n**Ultimate**:`bunx oh-my-openagent install`\n\n. - Already use Codex CLI? Choose\n**Light**:`npx lazycodex-ai install`\n\n. - Want one command without installing a host first? Choose\n**Senpi/native (beta)**:`npm i -g omo-ai@beta`\n\n.\n\nUltimate and Light are plugins that load into a host you already run. Senpi is standalone: it ships a pinned Senpi engine with OMO built in.\n\nFor Senpi, the `@beta`\n\ntag is required; bare `npm i -g omo-ai`\n\nfails by design. Do not install plain `omo`\n\nfrom npm: it is an unrelated package by a different author.\n\n**Strongly recommended: let an LLM agent install this for you.** The Ultimate edition setup involves subscription detection, model selection across 11 agents, and per-provider authentication — humans fat-finger these. An LLM agent reads the full guide and walks every step correctly.\n\nPaste this prompt into Claude Code, AmpCode, Cursor, or any agent:\n\n```\nInstall and configure oh-my-openagent by following the instructions here:\nhttps://raw.githubusercontent.com/code-yeongyu/oh-my-openagent/refs/heads/dev/docs/guide/installation.md\n```\n\nIf you only want the **Light edition** (Codex CLI), the installer asks whether to configure Codex for autonomous full-permissions mode. You can run it yourself in one line:\n\n```\nnpx lazycodex-ai install\n# non-interactive recommended mode:\nnpx lazycodex-ai install --no-tui --codex-autonomous\n```\n\nFor the Light edition, Bun is not required. Use `npx lazycodex-ai install`\n\nfrom a Node/npm environment. Global installation is not officially supported; the installer writes the Codex plugin into `~/.codex/`\n\n.\n\nFetch the full guide and follow it step by step:\n\n```\ncurl -fsSL https://raw.githubusercontent.com/code-yeongyu/oh-my-openagent/refs/heads/dev/docs/guide/installation.md\n```\n\nThe guide covers: platform selection, the subscription interview, provider authentication (Anthropic / Gemini / Copilot / Z.ai / OpenCode Zen), the agent-to-model matching matrix, modes (`ultrawork`\n\n, `team`\n\n, `hyperplan`\n\n), slash commands, the Light edition's Codex components, Team Mode, and uninstall. Don't summarize it; read it end to end.\n\nThe published npm package and CLI binary are still named `oh-my-opencode`\n\n(dual-published as `oh-my-openagent`\n\nduring the rename transition). Inside `opencode.json`\n\n, the compatibility layer prefers the plugin entry `oh-my-openagent`\n\n, while legacy `oh-my-opencode`\n\nentries still load with a warning. Runtime config is `~/.omo/omo.jsonc`\n\nplus walked project `.omo/omo.jsonc`\n\n. Legacy `oh-my-openagent.json[c]`\n\n/ `oh-my-opencode.json[c]`\n\nfiles are imported once by the migration engine and are not read afterward.\n\nThe recommended `bunx`\n\n/`npx`\n\ninvocation is `oh-my-openagent install`\n\n(or the original `oh-my-opencode install`\n\n). Once installed, the short command is `omo-agent-toolkit`\n\n. The `omo`\n\nbin was removed from these packages in this major release; the name now belongs to the senpi-native edition, installed with `npm i -g omo-ai@beta`\n\n(beta channel only). Do **not** use `bunx omo`\n\nor `npx omo`\n\n: `omo`\n\non npm is a different, unrelated package by a different author, and those commands resolve to it. The senpi edition's package name is `omo-ai`\n\n. `lazycodex-ai`\n\nis a single-purpose Node/npm installer package: `npx lazycodex-ai install`\n\nroutes directly to the Codex Light installer. It is not the Codex marketplace name (the marketplace repository is `code-yeongyu/lazycodex`\n\n). Codex sees marketplace `sisyphuslabs`\n\nand plugin `omo`\n\n, enabled as `omo@sisyphuslabs`\n\n.\n\nAnonymous telemetry is enabled by default to track active installations (DAU/WAU/MAU). For both products, a single event is sent **at most once per UTC day per machine** using a SHA256-hashed installation identifier (never the raw hostname), and PostHog person profiles are not created. The main plugin emits `omo_daily_active`\n\nfrom plugin load (`plugin_loaded`\n\n) and CLI run (`run_started`\n\n) sources; the Codex CLI Light edition emits `omo_codex_daily_active`\n\nfrom two sources (`install_completed`\n\nand `session_start`\n\n).\n\nOpt out per product:\n\n- Main plugin: set\n`\"telemetry\": false`\n\nin oh-my-openagent config,`OMO_DISABLE_POSTHOG=1`\n\n, or`OMO_SEND_ANONYMOUS_TELEMETRY=0`\n\n- Codex CLI Light edition:\n`OMO_CODEX_DISABLE_POSTHOG=1`\n\nor`OMO_CODEX_SEND_ANONYMOUS_TELEMETRY=0`\n\n(the global flags also disable Codex)\n\nSee [Privacy Policy](/code-yeongyu/oh-my-openagent/blob/dev/docs/legal/privacy-policy.md) and [Terms of Service](/code-yeongyu/oh-my-openagent/blob/dev/docs/legal/terms-of-service.md).\n\nWe're past the era of reading docs. Just paste this into your agent:\n\n```\nRead this and tell me why it's not just another boilerplate: https://raw.githubusercontent.com/code-yeongyu/oh-my-openagent/refs/heads/dev/README.md\n```\n\nWe are restructuring the codebase to support multiple agent harnesses (OpenCode, Codex, Pi, Claude Code, and others). The most urgent work is the package layering refactor: separating pure TypeScript core logic, MCP servers, skills, and adapter shims into distinct layers so the same logic can be reused across harnesses without duplication.\n\nIf you want to contribute, read the [ROADMAP](/code-yeongyu/oh-my-openagent/blob/dev/ROADMAP.md) first. PRs related to this refactor should use the `ROADMAP`\n\nlabel so we can track them.\n\nYou're actually reading this? Wild.\n\nInstall. Type `ultrawork`\n\n(or `ulw`\n\n). Done.\n\nEverything below, every feature, every optimization: you don't need to know any of it. It just works.\n\nEven with only the following subscriptions, `ultrawork`\n\nworks well (this project is not affiliated; these are personal recommendations):\n\n[ChatGPT Subscription ($20)](https://chatgpt.com/)[Kimi Code Subscription ($19)](https://www.kimi.com/code)[GLM Coding Plan ($10)](https://z.ai/subscribe)- If you're eligible for pay-per-token, using Kimi and Gemini models won't cost much.\n\n| Feature | Edition | What it does | |\n|---|---|---|---|\n| 🤖 | Discipline Agents |\nUltimate | Sisyphus orchestrates Hephaestus, Oracle, Librarian, Explore. A full AI dev team in parallel. |\n| 🧩 | Codex CLI Light Edition |\nLight | Portable OMO components (rules, comment-checker, git-bash, LSP, ultrawork, ulw-loop, ulw-execute continuation, telemetry, teammode, and more) running inside OpenAI Codex CLI. Install via `npx lazycodex-ai install` . |\n| 👥 | Team Mode (v4.0, opt-in) |\nUltimate | Lead agent + up to 8 parallel members, real-time tmux visualization, dedicated `team_*` tools. Powers `hyperplan` (5 hostile critics) and `security-research` (3 hunters + 2 PoC engineers).\n|\n| ⚡ | `ultrawork` / `ulw` |\nBoth | One word. Every agent activates. Doesn't stop until done. |\n| 🚪 |\n|\nUltimate | Analyzes true user intent before classifying or acting. No more literal misinterpretations. (Light edition only recognises the `ultrawork` /`ulw` keyword.) |\n| 🔗 | Hash-Anchored Edit Tool |\nUltimate | Hashline (`LINE#ID` ) edit/read tagging. Opt-in: set `hashline_edit: true` in `~/.omo/omo.jsonc` . Zero stale-line errors. Inspired by\n|\n| 🛠️ | LSP integration |\nBoth | Diagnostics, navigation, symbols, workspace rename. IDE precision for agents. Same LSP MCP server in both editions. |\n| 🔎 | AST-Grep |\nBoth | Pattern-aware code search and rewriting across 25 languages. Ultimate uses the MCP tools; Light uses the shared `ast-grep` skill with a provisioned `sg` . |\n| 🧠 | Background Agents |\nUltimate | Fire 5+ specialists in parallel. Context stays lean. Results when ready. |\n| 📚 | Built-in MCPs (web/docs/code search) |\nUltimate | Exa (web search), Context7 (official docs), Grep.app (GitHub search). Always on. |\n| 🔁 | Goal / `/goal` |\nUltimate | `/goal` sets a persistent thread objective. Idle continuation runs only when `goal.enabled` is true (default false). |\n| ✅ | Todo Enforcer |\nUltimate | Agent goes idle? System yanks it back. Your task gets done, period. |\n| 💬 | Comment Checker |\nBoth | No AI slop in comments. Code reads like a senior wrote it. |\n| 📐 | Rules Injection (`AGENTS.md` / `.omo/rules/**` ) |\nBoth | Project rules and AGENTS.md auto-loaded into the agent's context at every prompt. |\n| 🎯 | Ulw Loop |\nLight | Codex component + CLI for durable multi-goal orchestration with evidence audit. On OpenCode, use Goal / instead.`/goal` |\n| 🖥️ | Tmux Integration |\nUltimate | Full interactive terminal. REPLs, debuggers, TUIs. All live. |\n| 🔌 | Claude Code Compatible |\nUltimate | Your hooks, commands, skills, MCPs, and plugins? All work here. |\n| 🧬 | Skill-Embedded MCPs |\nUltimate | Skills carry their own MCP servers. No context bloat. |\n| 📋 | Prometheus Planner |\nUltimate | Interview-mode strategic planning before any execution. |\n| 🔍 | `/init-deep` |\nUltimate | Auto-generates hierarchical `AGENTS.md` files throughout your project. Great for both token efficiency and your agent's performance. |\n\nEdition legend.Ultimate= OpenCode-only (`bunx oh-my-openagent install`\n\n).Light= Codex CLI-only (`npx lazycodex-ai install`\n\n).Both= shipped in both editions, often with slightly different implementations under the hood.\n\n**Sisyphus** (`claude-opus-5`\n\n/ ** kimi-k3** /\n\n**/**\n\n`gpt-5.6-sol`\n\n**) is your main orchestrator. He plans, delegates to specialists, and drives tasks to completion with aggressive parallel execution. He does not stop halfway. Claude Opus 5 and Kimi K3 are the recommended defaults.**\n\n`glm-5.2`\n\n**Hephaestus** (`gpt-5.6-sol`\n\nthrough OpenAI, GitHub Copilot, Vercel, or OpenCode at medium effort) is your autonomous deep worker. Give him a goal, not a recipe. He explores the codebase, researches patterns, and executes end-to-end without hand-holding. *The Legitimate Craftsman.*\n\n**Prometheus** (`claude-fable-5`\n\n/ ** kimi-k3**) is your strategic planner. Interview mode: he asks questions, identifies scope, and builds a detailed plan before a single line of code is touched.\n\nEvery agent is tuned to its model's specific strengths. No manual model juggling. [Learn more →](/code-yeongyu/oh-my-openagent/blob/dev/docs/guide/overview.md)\n\nAnthropic\n\n[blocked OpenCode because of us.]That's why Hephaestus is called \"The Legitimate Craftsman.\" The irony is intentional.We run best on Opus or Kimi K3, but Kimi K3 + GPT-5.6 Sol already beats vanilla Claude Code. Zero config needed.\n\nOne agent is fast. A coordinated team is *devastating*.\n\n**Team Mode** turns oh-my-openagent from \"one agent with subagents\" into a real multi-agent system. A lead agent orchestrates a team of category-specialized members, all running **in parallel** and communicating through dedicated tools (`team_create`\n\n, `team_send_message`\n\n, `team_task_create`\n\n, `team_status`\n\n, ...). Watch every member work simultaneously in a tmux layout with focus + grid windows.\n\n```\n// ~/.omo/omo.jsonc  (or project .omo/omo.jsonc)\n{\n  \"team_mode\": {\n    \"enabled\": true,\n    \"max_parallel_members\": 4,\n    \"tmux_visualization\": true\n  }\n}\n```\n\nRestart opencode and the `team_*`\n\ntool family unlocks. Two skills already ride on top:\n\n— 5 hostile agents tear apart your plan from orthogonal angles before a single line of code is written.`hyperplan`\n\n— 3 vulnerability hunters + 2 PoC engineers audit your codebase in parallel, with severity calibrated by`security-research`\n\n*actual exploitability*.\n\nOff by default. Enable it when you want it.[Full Team Mode guide →]\n\nWhen Sisyphus delegates to a subagent, it doesn't pick a model. It picks a **category**. The category maps automatically to the right model:\n\n| Category | What it's for |\n|---|---|\n`visual-engineering` |\nFrontend, UI/UX, design |\n`deep` |\nAutonomous research + execution |\n`quick` |\nSingle-file changes, typos |\n`ultrabrain` |\nHard logic, architecture decisions |\n\nThe agent says what kind of work it needs; the harness picks the right model. `ultrabrain`\n\nroutes to GPT-5.6 Sol max (OpenAI / Vercel, then GitHub Copilot, then OpenCode). You touch nothing.\n\nYou dialed in your Claude Code setup. Good.\n\nEvery hook, command, skill, MCP, plugin works here unchanged. Full compatibility, including plugins.\n\nLSP, AST-Grep, Tmux, and MCP, actually integrated, not duct-taped together.\n\n**LSP**:`lsp_rename`\n\n,`lsp_goto_definition`\n\n,`lsp_find_references`\n\n,`lsp_diagnostics`\n\n. IDE precision for every agent.**AST-Grep**: Pattern-aware code search and rewriting across 25 languages.** Tmux**: Full interactive terminal. REPLs, debuggers, TUI apps. Your agent stays in session.** MCP**: Web search, official docs, GitHub code search. All baked in.\n\nMCP servers eat your context budget. We fixed that.\n\nSkills bring their own MCP servers. They spin up on demand, scoped to the task, and go away when done. The context window stays clean.\n\nThe harness problem is real. Most agent failures aren't the model's fault; it's the edit tool.\n\n\"None of these tools give the model a stable, verifiable identifier for the lines it wants to change... They all rely on the model reproducing content it already saw. When it can't - and it often can't - the user blames the model.\"\n\nInspired by [oh-my-pi](https://github.com/can1357/oh-my-pi), we built **Hashline**. When `hashline_edit`\n\nis enabled, every line the agent reads comes back tagged with a content hash:\n\n```\n11#VK| function hello() {\n22#XJ|   return \"world\";\n33#MB| }\n```\n\nThe agent edits by referencing those tags. If the file has changed since the last read, the hash won't match and the edit is rejected before any corruption. No whitespace reproduction. No stale-line errors.\n\nGrok Code Fast 1: **6.7% → 68.3%** success rate, just from changing the edit tool.\n\nRun `/init-deep`\n\n. It generates hierarchical `AGENTS.md`\n\nfiles:\n\n```\nproject/\n├── AGENTS.md              ← project-wide context\n├── src/\n│   ├── AGENTS.md          ← src-specific context\n│   └── components/\n│       └── AGENTS.md      ← component-specific context\n```\n\nAgents auto-read relevant context. Zero manual management.\n\nComplex task? Don't prompt and pray.\n\nPrometheus **interviews you like a real engineer**, identifies scope and ambiguities, and writes a verified plan to `.omo/plans/`\n\nbefore touching code. `/ulw-execute`\n\nthen starts an **Atlas** work session from that plan. The agent knows what it's building before it starts.\n\nSkills aren't just prompts. Each brings:\n\n- Domain-tuned system instructions.\n- Embedded MCP servers, on demand.\n- Scoped permissions so agents stay in bounds.\n\nBuilt-ins: `playwright`\n\n(browser automation), `git-master`\n\n(atomic commits, rebase surgery), `frontend`\n\n(design-first UI).\n\nAdd your own under `.opencode/skills/*/SKILL.md`\n\nor `~/.config/opencode/skills/*/SKILL.md`\n\n.\n\n**Want the full feature breakdown?** See the ** Features Documentation** for agents, hooks, tools, MCPs, and everything else in detail.\n\nNew to oh-my-openagent?Read theto understand what you have, or check the[Overview]for how agents collaborate.[Orchestration Guide]\n\nTo remove oh-my-openagent:\n\n-\n**Remove the plugin from your OpenCode config** Edit\n\n`~/.config/opencode/opencode.json`\n\n(or`opencode.jsonc`\n\n) and remove either`\"oh-my-openagent\"`\n\nor the legacy`\"oh-my-opencode\"`\n\nentry from the`plugin`\n\narray:\n\n```\n# Using jq\njq '.plugin = [.plugin[] | select(. != \"oh-my-openagent\" and . != \"oh-my-opencode\")]' \\\n    ~/.config/opencode/opencode.json > /tmp/oc.json && \\\n    mv /tmp/oc.json ~/.config/opencode/opencode.json\n```\n\n-\n**Remove configuration files (optional)**\n\n```\n# Remove the runtime config files\nrm -f ~/.omo/omo.jsonc ~/.omo/omo.json\n\n# Remove project config (if exists)\nrm -f .omo/omo.jsonc .omo/omo.json\n\n# Remove leftover legacy migration backups (if any)\nrm -f ~/.config/opencode/oh-my-openagent.jsonc ~/.config/opencode/oh-my-openagent.json \\\n      ~/.config/opencode/oh-my-opencode.jsonc ~/.config/opencode/oh-my-opencode.json \\\n      .opencode/oh-my-openagent.jsonc .opencode/oh-my-openagent.json \\\n      .opencode/oh-my-opencode.jsonc .opencode/oh-my-opencode.json\n```\n\n-\n**Verify removal**\n\n```\nopencode --version\n# Plugin should no longer be loaded\n```\n\n-\n**Remove omo-codex (Codex CLI Light edition)**\n\n```\nnpx lazycodex-ai uninstall\n# backward-compatible alias:\nnpx lazycodex-ai cleanup\n\nomo-agent-toolkit uninstall --platform=codex\n# backward-compatible alias:\nomo-agent-toolkit cleanup --platform=codex\n```\n\nThe uninstall command removes managed\n\n`sisyphuslabs`\n\nCodex cache/marketplace state, strips`omo@sisyphuslabs`\n\nplugin and hook-state blocks from`~/.codex/config.toml`\n\nafter writing a backup, and removes agent TOML links listed in the install manifest. If a specific project still has old project-local Codex plugin state, run the command from that project or pass`--project <path>`\n\n; it repairs known project-local`.codex/config.toml`\n\nconflicts and reports project-local`.codex`\n\nartifacts without deleting project-owned files.\n\nFeatures you'll think should've always existed. Once you use them, you can't go back.\n\nSee full [Features Documentation](/code-yeongyu/oh-my-openagent/blob/dev/docs/reference/features.md).\n\n**Quick Overview:**\n\n**Agents**: Sisyphus (the main agent), Prometheus (planner), Oracle (architecture/debugging), Librarian (docs/code search), Explore (fast codebase grep), Multimodal Looker**Background Agents**: Run multiple agents in parallel like a real dev team** LSP & AST Tools**: Refactoring, rename, diagnostics, AST-aware code search** Hash-anchored Edit Tool**(opt-in via`hashline_edit: true`\n\n):`LINE#ID`\n\nreferences validate content before applying every change. Surgical edits, zero stale-line errors**Context Injection**: Auto-inject AGENTS.md, README.md, conditional rules** Claude Code Compatibility**: Full hook system, commands, skills, agents, MCPs** Built-in MCPs**: websearch (Exa), context7 (docs), grep_app (GitHub search) — injected at runtime by the plugin; not visible in`opencode mcp list`\n\n(see[MCP docs](/code-yeongyu/oh-my-openagent/blob/dev/docs/reference/features.md#native-vs-plugin-injected-mcps))**Session Tools**: List, read, search, and analyze session history** Productivity Features**: Goal, Todo Enforcer, Comment Checker, Think Mode, and more** Doctor Command**: Built-in diagnostics (`bunx oh-my-opencode doctor`\n\n) verify plugin registration, config, models, and environment**Model Fallbacks**:`fallback_models`\n\ncan mix plain model strings with per-fallback object settings in the same array**File Prompts**: Load prompts from files with`file://`\n\nsupport in agent configurations**Session Recovery**: Automatic recovery from session errors, context window limits, and API failures** Model Setup**: Agent-model matching is built into the[Installation Guide](/code-yeongyu/oh-my-openagent/blob/dev/docs/guide/installation.md#step-5-understand-your-model-setup)\n\nOpinionated defaults, adjustable if you insist.\n\nSee [Configuration Documentation](/code-yeongyu/oh-my-openagent/blob/dev/docs/reference/configuration.md).\n\n**Quick Overview:**\n\n**Config Locations**: User`~/.omo/omo.jsonc`\n\nplus walked project`.omo/omo.jsonc`\n\nconfigs up to`$HOME`\n\n; closest wins. Legacy`oh-my-*`\n\nfiles are migrated once into`omo.jsonc`\n\n.**JSONC Support**: Comments and trailing commas supported** Agents**: Override models, temperatures, prompts, and permissions for any agent** Built-in Skills**:`playwright`\n\n(browser automation),`git-master`\n\n(atomic commits)**Sisyphus Agent**: Main orchestrator with Prometheus (Planner) and Metis (Plan Consultant)** Background Tasks**: Configure concurrency limits per provider/model** Categories**: Domain-specific task delegation (`visual-engineering`\n\n,`ultrabrain`\n\n,`deep`\n\n,`artistry`\n\n,`quick`\n\n,`unspecified-low`\n\n,`unspecified-high`\n\n,`writing`\n\n, plus custom names)**Hooks**: 54+ lifecycle hooks (61 with Team Mode), all configurable via`disabled_hooks`\n\n**MCPs**: Built-in websearch (Exa), context7 (docs), grep_app (GitHub search) — runtime-injected, not shown in`opencode mcp list`\n\n**LSP**: Full LSP support with refactoring tools** Experimental**: Aggressive truncation, auto-resume, and more\n\n**Want the philosophy?** Read the [Ultrawork Manifesto](/code-yeongyu/oh-my-openagent/blob/dev/docs/manifesto.md).\n\nI burned through $24K in LLM tokens on personal projects. Tried every tool. Configured everything to death. OpenCode won.\n\nEvery problem I hit, the fix is baked into this plugin. Install and go.\n\nIf OpenCode is Debian/Arch, oh-my-openagent is Ubuntu/[Omarchy](https://omarchy.org/).\n\nHeavily influenced by [AmpCode](https://ampcode.com) and [Claude Code](https://code.claude.com/docs/overview). Features ported, often improved. Still building. It's **Open** Code.\n\nOther harnesses promise multi-model orchestration. We ship it. Stability too. And features that actually work.\n\nI'm this project's most obsessive user:\n\n- Which model has the sharpest logic?\n- Who's the debugging god?\n- Who writes the best prose?\n- Who dominates frontend?\n- Who owns backend?\n- What's fastest for daily driving?\n- What are competitors shipping?\n\nThis plugin is the distillation. Take the best. Got improvements? PRs welcome.\n\n**Stop agonizing over harness choices.**\n**I'll research, steal the best, and ship it here.**\n\nSounds arrogant? Have a better way? Contribute. You're welcome.\n\nNo affiliation with any project or model mentioned. Just personal experimentation.\n\nCredit: The LazyCodex name idea is inspired by [LazyVim](https://github.com/LazyVim/LazyVim). The Ultragoal and UltraQA ideas are inspired by [oh-my-codex](https://github.com/Yeachan-Heo/oh-my-codex), reimplemented from concept for OmO.\n\n99% of this project was built with OpenCode. I don't really know TypeScript, **but I personally reviewed and largely rewrote this doc.**\n\n[Indent](https://indentcorp.com)- Makers of Spray (influencer marketing solution), vovushop (cross-border commerce platform), and vreview (AI commerce review marketing solution).\n\n[Google](https://google.com)[Microsoft](https://microsoft.com)[Vercel](https://vercel.com)[ELESTYLE](https://elestyle.jp)- Makers of elepay (multi-mobile payment gateway) and OneQR (mobile application SaaS for cashless solutions).\n\n[Deepgram](https://deepgram.com)\n\n*Special thanks to @junhoyeo for this amazing hero image.*", "url": "https://wpnews.pro/news/oh-my-openagent", "canonical_source": "https://github.com/code-yeongyu/oh-my-openagent/tree/dev", "published_at": "2026-09-02 18:32:54+00:00", "updated_at": "2026-09-02 18:52:57.866727+00:00", "lang": "en", "topics": ["ai-agents", "developer-tools", "ai-products"], "entities": ["oh-my-openagent", "Yeongyu Kim", "Jobdori", "Anthropic", "OpenAI", "Codex", "OpenCode", "LazyCodex"], "alternates": {"html": "https://wpnews.pro/news/oh-my-openagent", "markdown": "https://wpnews.pro/news/oh-my-openagent.md", "text": "https://wpnews.pro/news/oh-my-openagent.txt", "jsonld": "https://wpnews.pro/news/oh-my-openagent.jsonld"}}