{ "schemaVersion": "1.0", "item": { "slug": "joko-orchestrator", "name": "Joko Orchestrator", "source": "tencent", "type": "skill", "category": "AI 智能", "sourceUrl": "https://clawhub.ai/oyi77/joko-orchestrator", "canonicalUrl": "https://clawhub.ai/oyi77/joko-orchestrator", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/joko-orchestrator", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=joko-orchestrator", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "SKILL.md", "_meta.json" ], "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/joko-orchestrator" }, "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/joko-orchestrator", "agentPageUrl": "https://openagent3.xyz/skills/joko-orchestrator/agent", "manifestUrl": "https://openagent3.xyz/skills/joko-orchestrator/agent.json", "briefUrl": "https://openagent3.xyz/skills/joko-orchestrator/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": "Autonomous Skill Orchestrator v2.0", "body": "Inspired by oh-my-opencode's three-layer architecture, adapted for OpenClaw's ecosystem." }, { "title": "Core Philosophy", "body": "Traditional AI follows: user asks → AI responds. This fails for complex work because:\n\nContext overload: Large tasks exceed context windows\nCognitive drift: AI loses track mid-task\nVerification gaps: No systematic completeness check\nHuman bottleneck: Requires constant intervention\n\nThis skill solves these through specialization and delegation." }, { "title": "Architecture", "body": "┌─────────────────────────────────────────────────────────┐\n│ PLANNING LAYER (Interview + Plan Generation) │\n│ • Clarify intent through interview │\n│ • Generate structured work plan │\n│ • Review plan for gaps │\n└─────────────────────────────────────────────────────────┘\n ↓\n┌─────────────────────────────────────────────────────────┐\n│ ORCHESTRATION LAYER (Atlas - The Conductor) │\n│ • Read plan, delegate tasks │\n│ • Accumulate wisdom across tasks │\n│ • Verify results independently │\n│ • NEVER write code directly — only delegate │\n└─────────────────────────────────────────────────────────┘\n ↓\n┌─────────────────────────────────────────────────────────┐\n│ EXECUTION LAYER (Sub-agents via sessions_spawn) │\n│ • Focused task execution │\n│ • Return results + learnings │\n│ • Isolated context per task │\n└─────────────────────────────────────────────────────────┘" }, { "title": "Explicit Triggers", "body": "\"use autonomous-skill-orchestrator\"\n\"activate autonomous-skill-orchestrator\"\n\"start autonomous orchestration\"\n\"ulw\" or \"ultrawork\" (magic keyword mode)" }, { "title": "Magic Word: ultrawork / ulw", "body": "Include ultrawork or ulw in any prompt to activate full orchestration mode automatically.\nThe agent figures out the rest — parallel agents, background tasks, deep exploration, and relentless execution until completion." }, { "title": "Step 1.1: Interview", "body": "Before planning, gather clarity through brief interview:\n\nAsk only what's needed:\n\nWhat's the core objective?\nWhat are the boundaries (what's NOT in scope)?\nAny constraints or preferences?\nHow do we know when it's done?\n\nInterview Style by Intent:\n\nIntentFocusExample QuestionsRefactoringSafety\"What tests verify current behavior?\"Build NewPatterns\"Follow existing conventions or deviate?\"Debug/FixReproduction\"Steps to reproduce? Error messages?\"ResearchScope\"Depth vs breadth? Time constraints?\"" }, { "title": "Step 1.2: Plan Generation", "body": "After interview, generate structured plan:\n\n## Work Plan: [Title]\n\n### Objective\n[One sentence, frozen intent]\n\n### Tasks\n- [ ] Task 1: [Description]\n - Acceptance: [How to verify completion]\n - References: [Files, docs, skills needed]\n - Category: [quick|general|deep|creative]\n \n- [ ] Task 2: ...\n\n### Guardrails\n- MUST: [Required constraints]\n- MUST NOT: [Forbidden actions]\n\n### Verification\n[How to verify overall completion]" }, { "title": "Step 1.3: Plan Review (Self-Momus)", "body": "Before execution, validate:\n\nEach task has clear acceptance criteria\n References are concrete (not vague)\n No scope creep beyond objective\n Dependencies between tasks are explicit\n Guardrails are actionable\n\nIf any check fails, refine plan before proceeding." }, { "title": "Conductor Rules", "body": "The orchestrator:\n\n✅ CAN read files to understand context\n✅ CAN run commands to verify results\n✅ CAN search patterns with grep/glob\n✅ CAN spawn sub-agents for work\n\nThe orchestrator:\n\n❌ MUST NOT write/edit code directly\n❌ MUST NOT trust sub-agent claims blindly\n❌ MUST NOT skip verification" }, { "title": "Step 2.1: Task Delegation", "body": "Use sessions_spawn with category-appropriate configuration:\n\nCategoryUse ForModel HintTimeoutquickTrivial tasks, single file changesfast model2-5 mingeneralStandard implementationdefault5-10 mindeepComplex logic, architecturethinking model10-20 mincreativeUI/UX, content generationcreative model5-10 minresearchDocs, codebase explorationfast + broad5 min\n\nDelegation Template:\n\nsessions_spawn(\n label: \"task-{n}-{short-desc}\",\n task: \"\"\"\n ## Task\n {exact task from plan}\n \n ## Expected Outcome\n {acceptance criteria}\n \n ## Context\n {accumulated wisdom from previous tasks}\n \n ## Constraints\n - MUST: {guardrails}\n - MUST NOT: {forbidden actions}\n \n ## References\n {relevant files, docs}\n \"\"\",\n runTimeoutSeconds: {based on category}\n)" }, { "title": "Step 2.2: Parallel Execution", "body": "Identify independent tasks (no file conflicts, no dependencies) and spawn them simultaneously:\n\n# Tasks 2, 3, 4 have no dependencies\nsessions_spawn(label=\"task-2\", task=\"...\")\nsessions_spawn(label=\"task-3\", task=\"...\")\nsessions_spawn(label=\"task-4\", task=\"...\")\n# All run in parallel" }, { "title": "Step 2.3: Wisdom Accumulation", "body": "After each task completion, extract and record:\n\n## Wisdom Log\n\n### Conventions Discovered\n- [Pattern found in codebase]\n\n### Successful Approaches\n- [What worked]\n\n### Gotchas\n- [Pitfalls to avoid]\n\n### Commands Used\n- [Useful commands for similar tasks]\n\nStore in: memory/orchestrator-wisdom.md (append-only during session)\n\nPass accumulated wisdom to ALL subsequent sub-agents." }, { "title": "Step 2.4: Independent Verification", "body": "NEVER trust sub-agent claims. After each task:\n\nRead actual changed files\nRun tests/linting if applicable\nVerify acceptance criteria independently\nCross-reference with plan requirements\n\nIf verification fails:\n\nLog the failure in wisdom\nRe-delegate with failure context\nMax 2 retries per task, then escalate to user" }, { "title": "Step 3.1: Final Verification", "body": "All tasks marked complete\nAll acceptance criteria verified\nNo unresolved issues in wisdom log" }, { "title": "Step 3.2: Summary Report", "body": "## Orchestration Complete\n\n### Completed Tasks\n- [x] Task 1: {summary}\n- [x] Task 2: {summary}\n\n### Learnings\n{key wisdom accumulated}\n\n### Files Changed\n{list of modified files}\n\n### Next Steps (if any)\n{recommendations}" }, { "title": "Halt Conditions (Immediate Stop)", "body": "User issues explicit stop command\nIrreversible destructive action detected\nScope expansion beyond frozen intent\n3+ consecutive task failures\nSub-agent attempts to spawn further sub-agents (no recursion)" }, { "title": "Risk Classification", "body": "ClassDescriptionActionAIrreversible, destructive, or unboundedHALT immediatelyBBounded, resolvable with clarificationPause, ask userCCosmetic, non-operativeProceed with note" }, { "title": "Forbidden Actions", "body": "Creating new autonomous orchestrators\nModifying this skill file\nAccessing credentials without explicit need\nExternal API calls not in original scope\nRecursive spawning (sub-agents spawning sub-agents)" }, { "title": "Stop Commands", "body": "User can stop at any time with:\n\n\"stop\"\n\"halt\"\n\"cancel orchestration\"\n\"abort\"\n\nOn stop: immediately terminate all spawned sessions, output summary of completed work, await new instructions." }, { "title": "During Orchestration", "body": "Append to memory/orchestrator-wisdom.md for learnings\nReference existing memory files for context" }, { "title": "After Orchestration", "body": "Update daily memory with orchestration summary\nPersist significant learnings to MEMORY.md if valuable" }, { "title": "Example Usage", "body": "Simple (magic word):\n\nulw refactor the authentication module to use JWT\n\nExplicit activation:\n\nactivate autonomous-skill-orchestrator\n\nBuild a REST API with user registration, login, and profile endpoints\n\nWith constraints:\n\nuse autonomous-skill-orchestrator\n- Build payment integration with Stripe\n- MUST: Use existing database patterns\n- MUST NOT: Store card numbers locally\n- Deadline: Complete core flow only" } ], "body": "Autonomous Skill Orchestrator v2.0\n\nInspired by oh-my-opencode's three-layer architecture, adapted for OpenClaw's ecosystem.\n\nCore Philosophy\n\nTraditional AI follows: user asks → AI responds. This fails for complex work because:\n\nContext overload: Large tasks exceed context windows\nCognitive drift: AI loses track mid-task\nVerification gaps: No systematic completeness check\nHuman bottleneck: Requires constant intervention\n\nThis skill solves these through specialization and delegation.\n\nArchitecture\n┌─────────────────────────────────────────────────────────┐\n│ PLANNING LAYER (Interview + Plan Generation) │\n│ • Clarify intent through interview │\n│ • Generate structured work plan │\n│ • Review plan for gaps │\n└─────────────────────────────────────────────────────────┘\n ↓\n┌─────────────────────────────────────────────────────────┐\n│ ORCHESTRATION LAYER (Atlas - The Conductor) │\n│ • Read plan, delegate tasks │\n│ • Accumulate wisdom across tasks │\n│ • Verify results independently │\n│ • NEVER write code directly — only delegate │\n└─────────────────────────────────────────────────────────┘\n ↓\n┌─────────────────────────────────────────────────────────┐\n│ EXECUTION LAYER (Sub-agents via sessions_spawn) │\n│ • Focused task execution │\n│ • Return results + learnings │\n│ • Isolated context per task │\n└─────────────────────────────────────────────────────────┘\n\nActivation\nExplicit Triggers\n\"use autonomous-skill-orchestrator\"\n\"activate autonomous-skill-orchestrator\"\n\"start autonomous orchestration\"\n\"ulw\" or \"ultrawork\" (magic keyword mode)\nMagic Word: ultrawork / ulw\n\nInclude ultrawork or ulw in any prompt to activate full orchestration mode automatically. The agent figures out the rest — parallel agents, background tasks, deep exploration, and relentless execution until completion.\n\nPhase 1: Planning (Prometheus Mode)\nStep 1.1: Interview\n\nBefore planning, gather clarity through brief interview:\n\nAsk only what's needed:\n\nWhat's the core objective?\nWhat are the boundaries (what's NOT in scope)?\nAny constraints or preferences?\nHow do we know when it's done?\n\nInterview Style by Intent:\n\nIntent\tFocus\tExample Questions\nRefactoring\tSafety\t\"What tests verify current behavior?\"\nBuild New\tPatterns\t\"Follow existing conventions or deviate?\"\nDebug/Fix\tReproduction\t\"Steps to reproduce? Error messages?\"\nResearch\tScope\t\"Depth vs breadth? Time constraints?\"\nStep 1.2: Plan Generation\n\nAfter interview, generate structured plan:\n\n## Work Plan: [Title]\n\n### Objective\n[One sentence, frozen intent]\n\n### Tasks\n- [ ] Task 1: [Description]\n - Acceptance: [How to verify completion]\n - References: [Files, docs, skills needed]\n - Category: [quick|general|deep|creative]\n \n- [ ] Task 2: ...\n\n### Guardrails\n- MUST: [Required constraints]\n- MUST NOT: [Forbidden actions]\n\n### Verification\n[How to verify overall completion]\n\nStep 1.3: Plan Review (Self-Momus)\n\nBefore execution, validate:\n\n Each task has clear acceptance criteria\n References are concrete (not vague)\n No scope creep beyond objective\n Dependencies between tasks are explicit\n Guardrails are actionable\n\nIf any check fails, refine plan before proceeding.\n\nPhase 2: Orchestration (Atlas Mode)\nConductor Rules\n\nThe orchestrator:\n\n✅ CAN read files to understand context\n✅ CAN run commands to verify results\n✅ CAN search patterns with grep/glob\n✅ CAN spawn sub-agents for work\n\nThe orchestrator:\n\n❌ MUST NOT write/edit code directly\n❌ MUST NOT trust sub-agent claims blindly\n❌ MUST NOT skip verification\nStep 2.1: Task Delegation\n\nUse sessions_spawn with category-appropriate configuration:\n\nCategory\tUse For\tModel Hint\tTimeout\nquick\tTrivial tasks, single file changes\tfast model\t2-5 min\ngeneral\tStandard implementation\tdefault\t5-10 min\ndeep\tComplex logic, architecture\tthinking model\t10-20 min\ncreative\tUI/UX, content generation\tcreative model\t5-10 min\nresearch\tDocs, codebase exploration\tfast + broad\t5 min\n\nDelegation Template:\n\nsessions_spawn(\n label: \"task-{n}-{short-desc}\",\n task: \"\"\"\n ## Task\n {exact task from plan}\n \n ## Expected Outcome\n {acceptance criteria}\n \n ## Context\n {accumulated wisdom from previous tasks}\n \n ## Constraints\n - MUST: {guardrails}\n - MUST NOT: {forbidden actions}\n \n ## References\n {relevant files, docs}\n \"\"\",\n runTimeoutSeconds: {based on category}\n)\n\nStep 2.2: Parallel Execution\n\nIdentify independent tasks (no file conflicts, no dependencies) and spawn them simultaneously:\n\n# Tasks 2, 3, 4 have no dependencies\nsessions_spawn(label=\"task-2\", task=\"...\")\nsessions_spawn(label=\"task-3\", task=\"...\")\nsessions_spawn(label=\"task-4\", task=\"...\")\n# All run in parallel\n\nStep 2.3: Wisdom Accumulation\n\nAfter each task completion, extract and record:\n\n## Wisdom Log\n\n### Conventions Discovered\n- [Pattern found in codebase]\n\n### Successful Approaches\n- [What worked]\n\n### Gotchas\n- [Pitfalls to avoid]\n\n### Commands Used\n- [Useful commands for similar tasks]\n\n\nStore in: memory/orchestrator-wisdom.md (append-only during session)\n\nPass accumulated wisdom to ALL subsequent sub-agents.\n\nStep 2.4: Independent Verification\n\nNEVER trust sub-agent claims. After each task:\n\nRead actual changed files\nRun tests/linting if applicable\nVerify acceptance criteria independently\nCross-reference with plan requirements\n\nIf verification fails:\n\nLog the failure in wisdom\nRe-delegate with failure context\nMax 2 retries per task, then escalate to user\nPhase 3: Completion\nStep 3.1: Final Verification\nAll tasks marked complete\nAll acceptance criteria verified\nNo unresolved issues in wisdom log\nStep 3.2: Summary Report\n## Orchestration Complete\n\n### Completed Tasks\n- [x] Task 1: {summary}\n- [x] Task 2: {summary}\n\n### Learnings\n{key wisdom accumulated}\n\n### Files Changed\n{list of modified files}\n\n### Next Steps (if any)\n{recommendations}\n\nSafety Guardrails\nHalt Conditions (Immediate Stop)\nUser issues explicit stop command\nIrreversible destructive action detected\nScope expansion beyond frozen intent\n3+ consecutive task failures\nSub-agent attempts to spawn further sub-agents (no recursion)\nRisk Classification\nClass\tDescription\tAction\nA\tIrreversible, destructive, or unbounded\tHALT immediately\nB\tBounded, resolvable with clarification\tPause, ask user\nC\tCosmetic, non-operative\tProceed with note\nForbidden Actions\nCreating new autonomous orchestrators\nModifying this skill file\nAccessing credentials without explicit need\nExternal API calls not in original scope\nRecursive spawning (sub-agents spawning sub-agents)\nStop Commands\n\nUser can stop at any time with:\n\n\"stop\"\n\"halt\"\n\"cancel orchestration\"\n\"abort\"\n\nOn stop: immediately terminate all spawned sessions, output summary of completed work, await new instructions.\n\nMemory Integration\nDuring Orchestration\nAppend to memory/orchestrator-wisdom.md for learnings\nReference existing memory files for context\nAfter Orchestration\nUpdate daily memory with orchestration summary\nPersist significant learnings to MEMORY.md if valuable\nExample Usage\n\nSimple (magic word):\n\nulw refactor the authentication module to use JWT\n\n\nExplicit activation:\n\nactivate autonomous-skill-orchestrator\n\nBuild a REST API with user registration, login, and profile endpoints\n\n\nWith constraints:\n\nuse autonomous-skill-orchestrator\n- Build payment integration with Stripe\n- MUST: Use existing database patterns\n- MUST NOT: Store card numbers locally\n- Deadline: Complete core flow only" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/oyi77/joko-orchestrator", "publisherUrl": "https://clawhub.ai/oyi77/joko-orchestrator", "owner": "oyi77", "version": "2.0.0", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/joko-orchestrator", "downloadUrl": "https://openagent3.xyz/downloads/joko-orchestrator", "agentUrl": "https://openagent3.xyz/skills/joko-orchestrator/agent", "manifestUrl": "https://openagent3.xyz/skills/joko-orchestrator/agent.json", "briefUrl": "https://openagent3.xyz/skills/joko-orchestrator/agent.md" } }