{
  "schemaVersion": "1.0",
  "item": {
    "slug": "oh-my-opencode",
    "name": "Oh My OpenCode",
    "source": "tencent",
    "type": "skill",
    "category": "开发工具",
    "sourceUrl": "https://clawhub.ai/McOso/oh-my-opencode",
    "canonicalUrl": "https://clawhub.ai/McOso/oh-my-opencode",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/oh-my-opencode",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=oh-my-opencode",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "_meta.json",
      "SKILL.md",
      "scripts/run-ulw.sh",
      "scripts/doctor.sh",
      "references/troubleshooting.md",
      "references/configuration.md"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "Download the package from Yavira.",
        "Extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the extracted folder."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-30T16:55:25.780Z",
      "expiresAt": "2026-05-07T16:55:25.780Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
        "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null
      },
      "scope": "source",
      "summary": "Source download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this source.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/oh-my-opencode"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/oh-my-opencode",
    "agentPageUrl": "https://openagent3.xyz/skills/oh-my-opencode/agent",
    "manifestUrl": "https://openagent3.xyz/skills/oh-my-opencode/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/oh-my-opencode/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "Download the package from Yavira.",
      "Extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the extracted folder."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "Oh My OpenCode",
        "body": "Multi-agent orchestration plugin that transforms OpenCode into a full agent harness with specialized agents, background task execution, category-based model routing, and autonomous work modes.\n\nPackage: oh-my-opencode (install via bunx oh-my-opencode install)\nRepository: https://github.com/code-yeongyu/oh-my-opencode\nSchema: https://raw.githubusercontent.com/code-yeongyu/oh-my-opencode/master/assets/oh-my-opencode.schema.json"
      },
      {
        "title": "Prerequisites",
        "body": "OpenCode installed and configured (opencode --version should be 1.0.150+)\ncurl -fsSL https://opencode.ai/install | bash\n# or: npm install -g opencode-ai\n# or: bun install -g opencode-ai\n\n\nAt least one LLM provider authenticated (opencode auth login)\nStrongly recommended: Anthropic Claude Pro/Max subscription (Sisyphus uses Claude Opus 4.5)"
      },
      {
        "title": "Installation",
        "body": "Run the interactive installer:\n\nbunx oh-my-opencode install\n\nNon-interactive mode with provider flags:\n\nbunx oh-my-opencode install --no-tui \\\n  --claude=<yes|no|max20> \\\n  --openai=<yes|no> \\\n  --gemini=<yes|no> \\\n  --copilot=<yes|no> \\\n  --opencode-zen=<yes|no> \\\n  --zai-coding-plan=<yes|no>\n\nVerify:\n\nopencode --version\ncat ~/.config/opencode/opencode.json  # should contain \"oh-my-opencode\" in plugin array"
      },
      {
        "title": "Mode 1: Ultrawork (Quick Autonomous Work)",
        "body": "Include ultrawork or ulw in your prompt. That's it.\n\nulw add authentication to my Next.js app\n\nThe agent will automatically:\n\nExplore your codebase to understand existing patterns\nResearch best practices via specialized background agents\nImplement the feature following your conventions\nVerify with diagnostics and tests\nKeep working until 100% complete"
      },
      {
        "title": "Mode 2: Prometheus (Precise Planned Work)",
        "body": "For complex or critical tasks:\n\nPress Tab → switches to Prometheus (Planner) mode\nDescribe your work → Prometheus interviews you, asking clarifying questions while researching your codebase\nConfirm the plan → review generated plan in .sisyphus/plans/*.md\nRun /start-work → Atlas orchestrator takes over:\n\nDistributes tasks to specialized sub-agents\nVerifies each task completion independently\nAccumulates learnings across tasks\nTracks progress across sessions (resume anytime)\n\nCritical rule: Do NOT use Atlas without /start-work. Prometheus and Atlas are a pair — always use them together."
      },
      {
        "title": "Agents",
        "body": "All agents are enabled by default. Each has a default model and provider priority fallback chain.\n\nAgentRoleDefault ModelProvider Priority ChainSisyphusPrimary orchestratorclaude-opus-4-5anthropic → kimi-for-coding → zai-coding-plan → openai → googleSisyphus-JuniorFocused task executor (used by delegate_task with categories)Determined by categoryPer-category chainHephaestusAutonomous deep worker — goal-oriented, explores before actinggpt-5.2-codex (medium)openai → github-copilot → opencode (requires gpt-5.2-codex)OracleArchitecture, debugging, high-IQ reasoning (read-only)gpt-5.2openai → google → anthropicLibrarianOfficial docs, OSS search, remote codebase analysisglm-4.7zai-coding-plan → opencode → anthropicExploreFast codebase grep (contextual search)claude-haiku-4-5anthropic → github-copilot → opencodeMultimodal LookerImage/PDF/diagram analysisgemini-3-flashgoogle → openai → zai-coding-plan → kimi-for-coding → anthropic → opencodePrometheusWork planner (interview-based plan generation)claude-opus-4-5anthropic → kimi-for-coding → openai → googleMetisPre-planning consultant (ambiguity/failure-point analysis)claude-opus-4-5anthropic → kimi-for-coding → openai → googleMomusPlan reviewer (clarity, verifiability, completeness)gpt-5.2openai → anthropic → googleAtlasPlan orchestrator (executes Prometheus plans via /start-work)k2p5 / claude-sonnet-4-5kimi-for-coding → opencode → anthropic → openai → googleOpenCode-BuilderDefault build agent (disabled by default when Sisyphus is active)System defaultSystem default"
      },
      {
        "title": "Agent Invocation",
        "body": "Agents are invoked via delegate_task() or the --agent CLI flag — NOT with @ prefix.\n\n// Invoke a specific agent\ndelegate_task(subagent_type=\"oracle\", prompt=\"Review this architecture...\")\n\n// Invoke via category (routes to Sisyphus-Junior with category model)\ndelegate_task(category=\"visual-engineering\", load_skills=[\"frontend-ui-ux\"], prompt=\"...\")\n\n// Background execution (non-blocking)\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find auth patterns...\")\n\nCLI:\n\nopencode --agent oracle\nopencode run --agent librarian \"Explain how auth works in this codebase\""
      },
      {
        "title": "When to Use Which Agent",
        "body": "SituationAgentGeneral coding tasksSisyphus (default)Autonomous goal-oriented deep workHephaestus (requires gpt-5.2-codex)Architecture decisions, debugging after 2+ failuresOracleLooking up library docs, finding OSS examplesLibrarianFinding code patterns in your codebaseExploreAnalyzing images, PDFs, diagramsMultimodal LookerComplex multi-day projects needing a planPrometheus + Atlas (via Tab → /start-work)Pre-planning scope analysisMetisReviewing a generated plan for gapsMomusQuick single-file changesdelegate_task with quick category"
      },
      {
        "title": "Categories",
        "body": "Categories route tasks to Sisyphus-Junior with domain-optimized models via delegate_task().\n\nCategoryDefault ModelVariantProvider Priority ChainBest Forvisual-engineeringgemini-3-pro—google → anthropic → zai-coding-planFrontend, UI/UX, design, styling, animationultrabraingpt-5.2-codexxhighopenai → google → anthropicDeep logical reasoning, complex architecturedeepgpt-5.2-codexmediumopenai → anthropic → googleGoal-oriented autonomous problem-solving (Hephaestus-style)artistrygemini-3-promaxgoogle → anthropic → openaiCreative/novel approaches, unconventional solutionsquickclaude-haiku-4-5—anthropic → google → opencodeTrivial tasks, single file changes, typo fixesunspecified-lowclaude-sonnet-4-5—anthropic → openai → googleGeneral tasks, low effortunspecified-highclaude-opus-4-5maxanthropic → openai → googleGeneral tasks, high effortwritinggemini-3-flash—google → anthropic → zai-coding-plan → openaiDocumentation, prose, technical writing"
      },
      {
        "title": "Category Usage",
        "body": "delegate_task(category=\"visual-engineering\", load_skills=[\"frontend-ui-ux\"], prompt=\"Create a dashboard component\")\ndelegate_task(category=\"ultrabrain\", load_skills=[], prompt=\"Design the payment processing flow\")\ndelegate_task(category=\"quick\", load_skills=[\"git-master\"], prompt=\"Fix the typo in README.md\")\ndelegate_task(category=\"deep\", load_skills=[], prompt=\"Investigate and fix the memory leak in the worker pool\")"
      },
      {
        "title": "Critical: Model Resolution Priority",
        "body": "Categories do NOT use their built-in defaults unless configured. Resolution order:\n\nUser-configured model (in oh-my-opencode.json) — highest priority\nCategory's built-in default (if category is in config)\nSystem default model (from opencode.json) — fallback\n\nTo use optimal models, add categories to your config. See references/configuration.md."
      },
      {
        "title": "Built-in Skills",
        "body": "SkillPurposeUsageplaywrightBrowser automation via Playwright MCP (default browser engine)load_skills=[\"playwright\"]agent-browserVercel's agent-browser CLI with session managementSwitch via browser_automation_engine configgit-masterGit expert: atomic commits, rebase/squash, history searchload_skills=[\"git-master\"]frontend-ui-uxDesigner-turned-developer for stunning UI/UXload_skills=[\"frontend-ui-ux\"]\n\nSkills are injected into subagents via delegate_task(load_skills=[...])."
      },
      {
        "title": "Slash Commands",
        "body": "CommandDescription/init-deepInitialize hierarchical AGENTS.md knowledge base/start-workExecute a Prometheus plan with Atlas orchestrator/ralph-loopStart self-referential development loop until completion/ulw-loopStart ultrawork loop — continues until completion/cancel-ralphCancel active Ralph Loop/refactorIntelligent refactoring with LSP, AST-grep, architecture analysis, TDD/stop-continuationStop all continuation mechanisms (ralph loop, todo continuation, boulder)"
      },
      {
        "title": "Background Agents",
        "body": "Fire multiple agents in parallel for exploration and research:\n\n// Launch background agents (non-blocking)\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find auth patterns in codebase\")\ndelegate_task(subagent_type=\"librarian\", run_in_background=true, prompt=\"Find JWT best practices\")\n\n// Collect results when needed\nbackground_output(task_id=\"bg_abc123\")\n\n// Cancel all background tasks before final answer\nbackground_cancel(all=true)"
      },
      {
        "title": "Concurrency Configuration",
        "body": "{\n  \"background_task\": {\n    \"defaultConcurrency\": 5,\n    \"staleTimeoutMs\": 180000,\n    \"providerConcurrency\": { \"anthropic\": 3, \"google\": 10 },\n    \"modelConcurrency\": { \"anthropic/claude-opus-4-5\": 2 }\n  }\n}\n\nPriority: modelConcurrency > providerConcurrency > defaultConcurrency"
      },
      {
        "title": "Tmux Integration",
        "body": "Run background agents in separate tmux panes for visual multi-agent execution:\n\n{\n  \"tmux\": {\n    \"enabled\": true,\n    \"layout\": \"main-vertical\",\n    \"main_pane_size\": 60\n  }\n}\n\nRequires running OpenCode in server mode inside a tmux session:\n\ntmux new -s dev\nopencode --port 4096\n\nLayout options: main-vertical (default), main-horizontal, tiled, even-horizontal, even-vertical"
      },
      {
        "title": "Pattern 1: Explore + Librarian (Research Phase)",
        "body": "// Internal codebase search\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find how auth middleware is implemented\")\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find error handling patterns in the API layer\")\n\n// External documentation search\ndelegate_task(subagent_type=\"librarian\", run_in_background=true, prompt=\"Find official JWT documentation and security recommendations\")\n\n// Continue working immediately — collect results when needed"
      },
      {
        "title": "Pattern 2: Category-Based Delegation (Implementation Phase)",
        "body": "// Frontend work → visual-engineering category\ndelegate_task(category=\"visual-engineering\", load_skills=[\"frontend-ui-ux\"], prompt=\"Build the settings page\")\n\n// Quick fix → quick category\ndelegate_task(category=\"quick\", load_skills=[\"git-master\"], prompt=\"Create atomic commit for auth changes\")\n\n// Hard problem → ultrabrain category\ndelegate_task(category=\"ultrabrain\", load_skills=[], prompt=\"Design the caching invalidation strategy\")"
      },
      {
        "title": "Pattern 3: Session Continuity",
        "body": "// First delegation returns a session_id\nresult = delegate_task(category=\"quick\", load_skills=[\"git-master\"], prompt=\"Fix the type error\")\n// session_id: \"ses_abc123\"\n\n// Follow-up uses session_id to preserve full context\ndelegate_task(session_id=\"ses_abc123\", prompt=\"Also fix the related test file\")"
      },
      {
        "title": "Core Commands",
        "body": "opencode                           # Start TUI\nopencode --port 4096               # Start TUI with server mode (for tmux integration)\nopencode -c                        # Continue last session\nopencode -s <session-id>           # Continue specific session\nopencode --agent <agent-name>      # Start with specific agent\nopencode -m provider/model         # Start with specific model"
      },
      {
        "title": "Non-Interactive Mode",
        "body": "opencode run \"Explain closures in JavaScript\"\nopencode run --agent oracle \"Review this architecture\"\nopencode run -m openai/gpt-5.2 \"Complex reasoning task\"\nopencode run --format json \"Query\"    # Raw JSON output"
      },
      {
        "title": "Auth & Provider Management",
        "body": "opencode auth login                # Add/configure a provider\nopencode auth list                 # List authenticated providers\nopencode auth logout               # Remove a provider\nopencode models                    # List all available models\nopencode models anthropic          # List models for specific provider\nopencode models --refresh          # Refresh models cache"
      },
      {
        "title": "Session Management",
        "body": "opencode session list              # List all sessions\nopencode session list -n 10        # Last 10 sessions\nopencode export <session-id>       # Export session as JSON\nopencode import session.json       # Import session\nopencode stats                     # Token usage and cost statistics\nopencode stats --days 7            # Stats for last 7 days"
      },
      {
        "title": "Plugin & MCP Management",
        "body": "bunx oh-my-opencode install        # Install/configure oh-my-opencode\nbunx oh-my-opencode doctor         # Diagnose configuration issues\nopencode mcp list                  # List configured MCP servers\nopencode mcp add                   # Add an MCP server"
      },
      {
        "title": "Server Mode",
        "body": "opencode serve --port 4096         # Headless server\nopencode web --port 4096           # Server with web UI\nopencode attach http://localhost:4096  # Attach TUI to running server"
      },
      {
        "title": "Built-in MCPs",
        "body": "Oh My OpenCode includes these MCP servers out of the box:\n\nMCPToolPurposeExaweb_search_exaWeb search with clean LLM-ready contentContext7resolve-library-id, query-docsOfficial library/framework documentation lookupGrep.appsearchGitHubSearch real-world code examples from public GitHub repos"
      },
      {
        "title": "Hooks",
        "body": "All hooks are enabled by default. Disable specific hooks via disabled_hooks config:\n\nHookPurposetodo-continuation-enforcerForces agent to continue if it quits halfwaycontext-window-monitorMonitors and manages context window usagesession-recoveryRecovers sessions after crashessession-notificationNotifies on session eventscomment-checkerPrevents AI from adding excessive code commentsgrep-output-truncatorTruncates large grep outputstool-output-truncatorTruncates large tool outputsdirectory-agents-injectorInjects AGENTS.md from subdirectories (auto-disabled on OpenCode 1.1.37+)directory-readme-injectorInjects README.md contextempty-task-response-detectorDetects and handles empty task responsesthink-modeExtended thinking mode controlanthropic-context-window-limit-recoveryRecovers from Anthropic context limitsrules-injectorInjects project rulesbackground-notificationNotifies when background tasks completeauto-update-checkerChecks for oh-my-opencode updatesstartup-toastShows startup notification (sub-feature of auto-update-checker)keyword-detectorDetects keywords like ultrawork/ulw to trigger modesagent-usage-reminderReminds to use specialized agentsnon-interactive-envHandles non-interactive environmentsinteractive-bash-sessionManages interactive bash/tmux sessionscompaction-context-injectorInjects context during compactionthinking-block-validatorValidates thinking blocksclaude-code-hooksClaude Code compatibility hooksralph-loopRalph Loop continuation mechanismpreemptive-compactionTriggers compaction before context overflowauto-slash-commandAuto-triggers slash commandssisyphus-junior-notepadNotepad for Sisyphus-Junior subagentsedit-error-recoveryRecovers from edit errorsdelegate-task-retryRetries failed task delegationsprometheus-md-onlyEnforces Prometheus markdown-only outputstart-workHandles /start-work commandatlasAtlas orchestrator hook"
      },
      {
        "title": "Do",
        "body": "Use ulw for quick autonomous tasks — just include the keyword in your prompt\nUse Prometheus + /start-work for complex projects — interview-based planning leads to better outcomes\nConfigure categories for your providers — ensures optimal model selection instead of falling back to system default\nFire explore/librarian agents in parallel — always use run_in_background=true\nUse session continuity — pass session_id for follow-up interactions with the same subagent\nLet the agent delegate — Sisyphus is an orchestrator, not a solo implementer\nRun bunx oh-my-opencode doctor to diagnose issues"
      },
      {
        "title": "Don't",
        "body": "Don't use Atlas without /start-work — Atlas requires a Prometheus plan\nDon't manually specify models for every agent — the fallback chain handles this\nDon't disable todo-continuation-enforcer — it's what keeps the agent completing work\nDon't use Claude Haiku for Sisyphus — Opus 4.5 is strongly recommended\nDon't run explore/librarian synchronously — always background them"
      },
      {
        "title": "When to Use This Skill",
        "body": "Installing or configuring oh-my-opencode\nUnderstanding agent roles and delegation patterns\nTroubleshooting model resolution or provider issues\nSetting up tmux integration for visual multi-agent execution\nConfiguring categories for cost optimization\nUnderstanding the ultrawork vs Prometheus workflow choice"
      },
      {
        "title": "When NOT to Use This Skill",
        "body": "General OpenCode usage unrelated to oh-my-opencode plugin features\nProvider authentication issues (use opencode auth directly)\nOpenCode core configuration (use OpenCode docs at https://opencode.ai/docs/)"
      },
      {
        "title": "Rules for the Agent",
        "body": "Package name is oh-my-opencode — NOT @anthropics/opencode or any other name\nUse bunx (officially recommended) — not npx for oh-my-opencode CLI commands\nAgent invocation uses --agent flag or delegate_task() — NOT @agent prefix\nNever change model settings or disable features unless the user explicitly requests it\nSisyphus strongly recommends Opus 4.5 — using other models degrades the experience significantly\nCategories do NOT use built-in defaults unless configured — always verify with bunx oh-my-opencode doctor --verbose\nPrometheus and Atlas are always paired — never use Atlas without a Prometheus plan\nBackground agents should always use run_in_background=true — never block on exploration\nSession IDs should be preserved and reused — saves 70%+ tokens on follow-ups\nWhen using Ollama, set stream: false — required to avoid JSON parse errors"
      },
      {
        "title": "Auto-Notify on Completion",
        "body": "Background tasks automatically notify when complete via the background-notification hook. No polling needed — the system pushes completion events. Use background_output(task_id=\"...\") only when you need to read the result."
      },
      {
        "title": "Reference Documents",
        "body": "Configuration Reference — Complete config with all agents, categories, provider chains, hooks, and options\nTroubleshooting Guide — Common issues and solutions"
      }
    ],
    "body": "Oh My OpenCode\n\nMulti-agent orchestration plugin that transforms OpenCode into a full agent harness with specialized agents, background task execution, category-based model routing, and autonomous work modes.\n\nPackage: oh-my-opencode (install via bunx oh-my-opencode install) Repository: https://github.com/code-yeongyu/oh-my-opencode Schema: https://raw.githubusercontent.com/code-yeongyu/oh-my-opencode/master/assets/oh-my-opencode.schema.json\n\nPrerequisites\nOpenCode installed and configured (opencode --version should be 1.0.150+)\ncurl -fsSL https://opencode.ai/install | bash\n# or: npm install -g opencode-ai\n# or: bun install -g opencode-ai\n\nAt least one LLM provider authenticated (opencode auth login)\nStrongly recommended: Anthropic Claude Pro/Max subscription (Sisyphus uses Claude Opus 4.5)\nInstallation\n\nRun the interactive installer:\n\nbunx oh-my-opencode install\n\n\nNon-interactive mode with provider flags:\n\nbunx oh-my-opencode install --no-tui \\\n  --claude=<yes|no|max20> \\\n  --openai=<yes|no> \\\n  --gemini=<yes|no> \\\n  --copilot=<yes|no> \\\n  --opencode-zen=<yes|no> \\\n  --zai-coding-plan=<yes|no>\n\n\nVerify:\n\nopencode --version\ncat ~/.config/opencode/opencode.json  # should contain \"oh-my-opencode\" in plugin array\n\nTwo Workflow Modes\nMode 1: Ultrawork (Quick Autonomous Work)\n\nInclude ultrawork or ulw in your prompt. That's it.\n\nulw add authentication to my Next.js app\n\n\nThe agent will automatically:\n\nExplore your codebase to understand existing patterns\nResearch best practices via specialized background agents\nImplement the feature following your conventions\nVerify with diagnostics and tests\nKeep working until 100% complete\nMode 2: Prometheus (Precise Planned Work)\n\nFor complex or critical tasks:\n\nPress Tab → switches to Prometheus (Planner) mode\nDescribe your work → Prometheus interviews you, asking clarifying questions while researching your codebase\nConfirm the plan → review generated plan in .sisyphus/plans/*.md\nRun /start-work → Atlas orchestrator takes over:\nDistributes tasks to specialized sub-agents\nVerifies each task completion independently\nAccumulates learnings across tasks\nTracks progress across sessions (resume anytime)\n\nCritical rule: Do NOT use Atlas without /start-work. Prometheus and Atlas are a pair — always use them together.\n\nAgents\n\nAll agents are enabled by default. Each has a default model and provider priority fallback chain.\n\nAgent\tRole\tDefault Model\tProvider Priority Chain\nSisyphus\tPrimary orchestrator\tclaude-opus-4-5\tanthropic → kimi-for-coding → zai-coding-plan → openai → google\nSisyphus-Junior\tFocused task executor (used by delegate_task with categories)\tDetermined by category\tPer-category chain\nHephaestus\tAutonomous deep worker — goal-oriented, explores before acting\tgpt-5.2-codex (medium)\topenai → github-copilot → opencode (requires gpt-5.2-codex)\nOracle\tArchitecture, debugging, high-IQ reasoning (read-only)\tgpt-5.2\topenai → google → anthropic\nLibrarian\tOfficial docs, OSS search, remote codebase analysis\tglm-4.7\tzai-coding-plan → opencode → anthropic\nExplore\tFast codebase grep (contextual search)\tclaude-haiku-4-5\tanthropic → github-copilot → opencode\nMultimodal Looker\tImage/PDF/diagram analysis\tgemini-3-flash\tgoogle → openai → zai-coding-plan → kimi-for-coding → anthropic → opencode\nPrometheus\tWork planner (interview-based plan generation)\tclaude-opus-4-5\tanthropic → kimi-for-coding → openai → google\nMetis\tPre-planning consultant (ambiguity/failure-point analysis)\tclaude-opus-4-5\tanthropic → kimi-for-coding → openai → google\nMomus\tPlan reviewer (clarity, verifiability, completeness)\tgpt-5.2\topenai → anthropic → google\nAtlas\tPlan orchestrator (executes Prometheus plans via /start-work)\tk2p5 / claude-sonnet-4-5\tkimi-for-coding → opencode → anthropic → openai → google\nOpenCode-Builder\tDefault build agent (disabled by default when Sisyphus is active)\tSystem default\tSystem default\nAgent Invocation\n\nAgents are invoked via delegate_task() or the --agent CLI flag — NOT with @ prefix.\n\n// Invoke a specific agent\ndelegate_task(subagent_type=\"oracle\", prompt=\"Review this architecture...\")\n\n// Invoke via category (routes to Sisyphus-Junior with category model)\ndelegate_task(category=\"visual-engineering\", load_skills=[\"frontend-ui-ux\"], prompt=\"...\")\n\n// Background execution (non-blocking)\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find auth patterns...\")\n\n\nCLI:\n\nopencode --agent oracle\nopencode run --agent librarian \"Explain how auth works in this codebase\"\n\nWhen to Use Which Agent\nSituation\tAgent\nGeneral coding tasks\tSisyphus (default)\nAutonomous goal-oriented deep work\tHephaestus (requires gpt-5.2-codex)\nArchitecture decisions, debugging after 2+ failures\tOracle\nLooking up library docs, finding OSS examples\tLibrarian\nFinding code patterns in your codebase\tExplore\nAnalyzing images, PDFs, diagrams\tMultimodal Looker\nComplex multi-day projects needing a plan\tPrometheus + Atlas (via Tab → /start-work)\nPre-planning scope analysis\tMetis\nReviewing a generated plan for gaps\tMomus\nQuick single-file changes\tdelegate_task with quick category\nCategories\n\nCategories route tasks to Sisyphus-Junior with domain-optimized models via delegate_task().\n\nCategory\tDefault Model\tVariant\tProvider Priority Chain\tBest For\nvisual-engineering\tgemini-3-pro\t—\tgoogle → anthropic → zai-coding-plan\tFrontend, UI/UX, design, styling, animation\nultrabrain\tgpt-5.2-codex\txhigh\topenai → google → anthropic\tDeep logical reasoning, complex architecture\ndeep\tgpt-5.2-codex\tmedium\topenai → anthropic → google\tGoal-oriented autonomous problem-solving (Hephaestus-style)\nartistry\tgemini-3-pro\tmax\tgoogle → anthropic → openai\tCreative/novel approaches, unconventional solutions\nquick\tclaude-haiku-4-5\t—\tanthropic → google → opencode\tTrivial tasks, single file changes, typo fixes\nunspecified-low\tclaude-sonnet-4-5\t—\tanthropic → openai → google\tGeneral tasks, low effort\nunspecified-high\tclaude-opus-4-5\tmax\tanthropic → openai → google\tGeneral tasks, high effort\nwriting\tgemini-3-flash\t—\tgoogle → anthropic → zai-coding-plan → openai\tDocumentation, prose, technical writing\nCategory Usage\ndelegate_task(category=\"visual-engineering\", load_skills=[\"frontend-ui-ux\"], prompt=\"Create a dashboard component\")\ndelegate_task(category=\"ultrabrain\", load_skills=[], prompt=\"Design the payment processing flow\")\ndelegate_task(category=\"quick\", load_skills=[\"git-master\"], prompt=\"Fix the typo in README.md\")\ndelegate_task(category=\"deep\", load_skills=[], prompt=\"Investigate and fix the memory leak in the worker pool\")\n\nCritical: Model Resolution Priority\n\nCategories do NOT use their built-in defaults unless configured. Resolution order:\n\nUser-configured model (in oh-my-opencode.json) — highest priority\nCategory's built-in default (if category is in config)\nSystem default model (from opencode.json) — fallback\n\nTo use optimal models, add categories to your config. See references/configuration.md.\n\nBuilt-in Skills\nSkill\tPurpose\tUsage\nplaywright\tBrowser automation via Playwright MCP (default browser engine)\tload_skills=[\"playwright\"]\nagent-browser\tVercel's agent-browser CLI with session management\tSwitch via browser_automation_engine config\ngit-master\tGit expert: atomic commits, rebase/squash, history search\tload_skills=[\"git-master\"]\nfrontend-ui-ux\tDesigner-turned-developer for stunning UI/UX\tload_skills=[\"frontend-ui-ux\"]\n\nSkills are injected into subagents via delegate_task(load_skills=[...]).\n\nSlash Commands\nCommand\tDescription\n/init-deep\tInitialize hierarchical AGENTS.md knowledge base\n/start-work\tExecute a Prometheus plan with Atlas orchestrator\n/ralph-loop\tStart self-referential development loop until completion\n/ulw-loop\tStart ultrawork loop — continues until completion\n/cancel-ralph\tCancel active Ralph Loop\n/refactor\tIntelligent refactoring with LSP, AST-grep, architecture analysis, TDD\n/stop-continuation\tStop all continuation mechanisms (ralph loop, todo continuation, boulder)\nProcess Management\nBackground Agents\n\nFire multiple agents in parallel for exploration and research:\n\n// Launch background agents (non-blocking)\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find auth patterns in codebase\")\ndelegate_task(subagent_type=\"librarian\", run_in_background=true, prompt=\"Find JWT best practices\")\n\n// Collect results when needed\nbackground_output(task_id=\"bg_abc123\")\n\n// Cancel all background tasks before final answer\nbackground_cancel(all=true)\n\nConcurrency Configuration\n{\n  \"background_task\": {\n    \"defaultConcurrency\": 5,\n    \"staleTimeoutMs\": 180000,\n    \"providerConcurrency\": { \"anthropic\": 3, \"google\": 10 },\n    \"modelConcurrency\": { \"anthropic/claude-opus-4-5\": 2 }\n  }\n}\n\n\nPriority: modelConcurrency > providerConcurrency > defaultConcurrency\n\nTmux Integration\n\nRun background agents in separate tmux panes for visual multi-agent execution:\n\n{\n  \"tmux\": {\n    \"enabled\": true,\n    \"layout\": \"main-vertical\",\n    \"main_pane_size\": 60\n  }\n}\n\n\nRequires running OpenCode in server mode inside a tmux session:\n\ntmux new -s dev\nopencode --port 4096\n\n\nLayout options: main-vertical (default), main-horizontal, tiled, even-horizontal, even-vertical\n\nParallel Execution Patterns\nPattern 1: Explore + Librarian (Research Phase)\n// Internal codebase search\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find how auth middleware is implemented\")\ndelegate_task(subagent_type=\"explore\", run_in_background=true, prompt=\"Find error handling patterns in the API layer\")\n\n// External documentation search\ndelegate_task(subagent_type=\"librarian\", run_in_background=true, prompt=\"Find official JWT documentation and security recommendations\")\n\n// Continue working immediately — collect results when needed\n\nPattern 2: Category-Based Delegation (Implementation Phase)\n// Frontend work → visual-engineering category\ndelegate_task(category=\"visual-engineering\", load_skills=[\"frontend-ui-ux\"], prompt=\"Build the settings page\")\n\n// Quick fix → quick category\ndelegate_task(category=\"quick\", load_skills=[\"git-master\"], prompt=\"Create atomic commit for auth changes\")\n\n// Hard problem → ultrabrain category\ndelegate_task(category=\"ultrabrain\", load_skills=[], prompt=\"Design the caching invalidation strategy\")\n\nPattern 3: Session Continuity\n// First delegation returns a session_id\nresult = delegate_task(category=\"quick\", load_skills=[\"git-master\"], prompt=\"Fix the type error\")\n// session_id: \"ses_abc123\"\n\n// Follow-up uses session_id to preserve full context\ndelegate_task(session_id=\"ses_abc123\", prompt=\"Also fix the related test file\")\n\nCLI Reference\nCore Commands\nopencode                           # Start TUI\nopencode --port 4096               # Start TUI with server mode (for tmux integration)\nopencode -c                        # Continue last session\nopencode -s <session-id>           # Continue specific session\nopencode --agent <agent-name>      # Start with specific agent\nopencode -m provider/model         # Start with specific model\n\nNon-Interactive Mode\nopencode run \"Explain closures in JavaScript\"\nopencode run --agent oracle \"Review this architecture\"\nopencode run -m openai/gpt-5.2 \"Complex reasoning task\"\nopencode run --format json \"Query\"    # Raw JSON output\n\nAuth & Provider Management\nopencode auth login                # Add/configure a provider\nopencode auth list                 # List authenticated providers\nopencode auth logout               # Remove a provider\nopencode models                    # List all available models\nopencode models anthropic          # List models for specific provider\nopencode models --refresh          # Refresh models cache\n\nSession Management\nopencode session list              # List all sessions\nopencode session list -n 10        # Last 10 sessions\nopencode export <session-id>       # Export session as JSON\nopencode import session.json       # Import session\nopencode stats                     # Token usage and cost statistics\nopencode stats --days 7            # Stats for last 7 days\n\nPlugin & MCP Management\nbunx oh-my-opencode install        # Install/configure oh-my-opencode\nbunx oh-my-opencode doctor         # Diagnose configuration issues\nopencode mcp list                  # List configured MCP servers\nopencode mcp add                   # Add an MCP server\n\nServer Mode\nopencode serve --port 4096         # Headless server\nopencode web --port 4096           # Server with web UI\nopencode attach http://localhost:4096  # Attach TUI to running server\n\nBuilt-in MCPs\n\nOh My OpenCode includes these MCP servers out of the box:\n\nMCP\tTool\tPurpose\nExa\tweb_search_exa\tWeb search with clean LLM-ready content\nContext7\tresolve-library-id, query-docs\tOfficial library/framework documentation lookup\nGrep.app\tsearchGitHub\tSearch real-world code examples from public GitHub repos\nHooks\n\nAll hooks are enabled by default. Disable specific hooks via disabled_hooks config:\n\nHook\tPurpose\ntodo-continuation-enforcer\tForces agent to continue if it quits halfway\ncontext-window-monitor\tMonitors and manages context window usage\nsession-recovery\tRecovers sessions after crashes\nsession-notification\tNotifies on session events\ncomment-checker\tPrevents AI from adding excessive code comments\ngrep-output-truncator\tTruncates large grep outputs\ntool-output-truncator\tTruncates large tool outputs\ndirectory-agents-injector\tInjects AGENTS.md from subdirectories (auto-disabled on OpenCode 1.1.37+)\ndirectory-readme-injector\tInjects README.md context\nempty-task-response-detector\tDetects and handles empty task responses\nthink-mode\tExtended thinking mode control\nanthropic-context-window-limit-recovery\tRecovers from Anthropic context limits\nrules-injector\tInjects project rules\nbackground-notification\tNotifies when background tasks complete\nauto-update-checker\tChecks for oh-my-opencode updates\nstartup-toast\tShows startup notification (sub-feature of auto-update-checker)\nkeyword-detector\tDetects keywords like ultrawork/ulw to trigger modes\nagent-usage-reminder\tReminds to use specialized agents\nnon-interactive-env\tHandles non-interactive environments\ninteractive-bash-session\tManages interactive bash/tmux sessions\ncompaction-context-injector\tInjects context during compaction\nthinking-block-validator\tValidates thinking blocks\nclaude-code-hooks\tClaude Code compatibility hooks\nralph-loop\tRalph Loop continuation mechanism\npreemptive-compaction\tTriggers compaction before context overflow\nauto-slash-command\tAuto-triggers slash commands\nsisyphus-junior-notepad\tNotepad for Sisyphus-Junior subagents\nedit-error-recovery\tRecovers from edit errors\ndelegate-task-retry\tRetries failed task delegations\nprometheus-md-only\tEnforces Prometheus markdown-only output\nstart-work\tHandles /start-work command\natlas\tAtlas orchestrator hook\nBest Practices\nDo\nUse ulw for quick autonomous tasks — just include the keyword in your prompt\nUse Prometheus + /start-work for complex projects — interview-based planning leads to better outcomes\nConfigure categories for your providers — ensures optimal model selection instead of falling back to system default\nFire explore/librarian agents in parallel — always use run_in_background=true\nUse session continuity — pass session_id for follow-up interactions with the same subagent\nLet the agent delegate — Sisyphus is an orchestrator, not a solo implementer\nRun bunx oh-my-opencode doctor to diagnose issues\nDon't\nDon't use Atlas without /start-work — Atlas requires a Prometheus plan\nDon't manually specify models for every agent — the fallback chain handles this\nDon't disable todo-continuation-enforcer — it's what keeps the agent completing work\nDon't use Claude Haiku for Sisyphus — Opus 4.5 is strongly recommended\nDon't run explore/librarian synchronously — always background them\nWhen to Use This Skill\nInstalling or configuring oh-my-opencode\nUnderstanding agent roles and delegation patterns\nTroubleshooting model resolution or provider issues\nSetting up tmux integration for visual multi-agent execution\nConfiguring categories for cost optimization\nUnderstanding the ultrawork vs Prometheus workflow choice\nWhen NOT to Use This Skill\nGeneral OpenCode usage unrelated to oh-my-opencode plugin features\nProvider authentication issues (use opencode auth directly)\nOpenCode core configuration (use OpenCode docs at https://opencode.ai/docs/)\nRules for the Agent\nPackage name is oh-my-opencode — NOT @anthropics/opencode or any other name\nUse bunx (officially recommended) — not npx for oh-my-opencode CLI commands\nAgent invocation uses --agent flag or delegate_task() — NOT @agent prefix\nNever change model settings or disable features unless the user explicitly requests it\nSisyphus strongly recommends Opus 4.5 — using other models degrades the experience significantly\nCategories do NOT use built-in defaults unless configured — always verify with bunx oh-my-opencode doctor --verbose\nPrometheus and Atlas are always paired — never use Atlas without a Prometheus plan\nBackground agents should always use run_in_background=true — never block on exploration\nSession IDs should be preserved and reused — saves 70%+ tokens on follow-ups\nWhen using Ollama, set stream: false — required to avoid JSON parse errors\nAuto-Notify on Completion\n\nBackground tasks automatically notify when complete via the background-notification hook. No polling needed — the system pushes completion events. Use background_output(task_id=\"...\") only when you need to read the result.\n\nReference Documents\nConfiguration Reference — Complete config with all agents, categories, provider chains, hooks, and options\nTroubleshooting Guide — Common issues and solutions"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/McOso/oh-my-opencode",
    "publisherUrl": "https://clawhub.ai/McOso/oh-my-opencode",
    "owner": "McOso",
    "version": "1.0.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/oh-my-opencode",
    "downloadUrl": "https://openagent3.xyz/downloads/oh-my-opencode",
    "agentUrl": "https://openagent3.xyz/skills/oh-my-opencode/agent",
    "manifestUrl": "https://openagent3.xyz/skills/oh-my-opencode/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/oh-my-opencode/agent.md"
  }
}