{ "schemaVersion": "1.0", "item": { "slug": "swarm-coding-skill", "name": "Hackathon Swarm Coding", "source": "tencent", "type": "skill", "category": "开发工具", "sourceUrl": "https://clawhub.ai/arunnadarasa/swarm-coding-skill", "canonicalUrl": "https://clawhub.ai/arunnadarasa/swarm-coding-skill", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/swarm-coding-skill", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=swarm-coding-skill", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "_meta.json", "README.md", "package.json", "SKILL.md", "orchestrator.js" ], "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. Then review README.md for any prerequisites, environment setup, or post-install checks. 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. Then review README.md for any prerequisites, environment setup, or post-install checks. 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/swarm-coding-skill" }, "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/swarm-coding-skill", "agentPageUrl": "https://openagent3.xyz/skills/swarm-coding-skill/agent", "manifestUrl": "https://openagent3.xyz/skills/swarm-coding-skill/agent.json", "briefUrl": "https://openagent3.xyz/skills/swarm-coding-skill/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. Then review README.md for any prerequisites, environment setup, or post-install checks. 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. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "documentation": { "source": "clawhub", "primaryDoc": "SKILL.md", "sections": [ { "title": "Swarm Coding Skill", "body": "Fully autonomous multi-agent software development. Given a plain-English prompt, the swarm designs, implements, tests, and delivers a complete project end-to-end.\n\nCore capability: Code generation via OpenRouter's qwen3-coder model. The orchestrator drives a Planner to create a manifest, then executes specialized worker roles (BackendDev, FrontendDev, QA, DevOps, etc.) in dependency order. All code is written to files; no interactive sessions.\n\nImportant: This skill generates code for review and deployment by the user. It does not make business decisions or operate autonomously in production. The user remains responsible for security, compliance, and operational decisions." }, { "title": "How It Works", "body": "Orchestrator (Planner role) analyzes your prompt, decides tech stack and architecture, and creates a swarm.yaml manifest with tasks and dependencies.\nWorker agents (BackendDev, FrontendDev, QA, DevOps) are spawned as sub-sessions. Each has a clear persona and works on its assigned files in a shared workspace.\nCoordination: The orchestrator tracks task completion and dependencies. When a task finishes, it marks it done and starts any unblocked downstream tasks.\nConflict avoidance: Files are partitioned by role (Backend owns server/, Frontend owns client/, etc.). If two roles need the same file, the manifest assigns an owner.\nQuality gates: QA must pass tests before integration; DevOps ensures containerization; no merge without green tests.\nDeliverable: You get a complete project directory with README, tests, Dockerfile, and optionally a GitHub repo or zip." }, { "title": "Usage", "body": "# In your main OpenClaw session, invoke:\n/trigger swarm-code \"Build a dashboard that shows Moltbook stats and ClawCredit status\"\n\nThe skill will:\n\nSpawn the orchestrator in an isolated session\nOrchestrator spawns workers sequentially or in parallel (based on dependencies)\nOutput a final summary and path to the completed project" }, { "title": "Requirements", "body": "Node.js v18+\nEnvironment variables (in .env at workspace root):\n\nRequired: OPENROUTER_API_KEY — OpenRouter API key with qwen/qwen3-coder access\nOptional: OPENROUTER_MODEL (default: qwen/qwen3-coder), MOCK=1 for dry-run\n\n\nInternet access for OpenRouter API (and optionally GitHub/Docker if deployment requested)\n\nImportant: The orchestrator reads .env from the workspace root (parent directory of this skill) and writes project files to swarm-projects/ and logs to .learnings/ in that same workspace root. Run in an isolated workspace to avoid exposing unrelated secrets." }, { "title": "Configuration", "body": "Store your OpenRouter key in .env at the workspace root:\n\nOPENROUTER_API_KEY=sk-or-...\n\nOptional overrides:\n\nOPENROUTER_MODEL=qwen/qwen3-coder\nMOCK=1 # dry-run, no API calls\n\nThe skill uses qwen/qwen3-coder by default. Ensure your OpenRouter key has that model enabled." }, { "title": "Output", "body": "The created project lives in swarm-projects// and includes:\n\nREADME.md with run instructions\npackage.json (or equivalent)\nSource code organized by component\ntest/ directory with automated tests\nDockerfile and docker-compose.yml (if applicable)\nCI/ with GitHub Actions workflow (optional)\nDECISIONS.md — Project memory documenting key architectural and technical decisions with rationale\n.learnings/ — Learning logs capturing errors, insights, and feature requests\n\nERRORS.md — Failures, exceptions, and recovery actions\nLEARNINGS.md — Corrections, better approaches, knowledge gaps\nFEATURE_REQUESTS.md — Requested capabilities that don't exist yet\n\n\nSWARM_SUMMARY.md — Execution summary with role performance, statistics, and next steps" }, { "title": "Continuous Improvement", "body": "The swarm skill automatically captures learnings during execution to improve future runs:" }, { "title": "What Gets Logged", "body": "Worker failures → .learnings/ERRORS.md with context and recovery suggestions\nBetter approaches discovered → .learnings/LEARNINGS.md (e.g., \"Simplified X by using Y\")\nUser corrections → .learnings/LEARNINGS.md when you override a decision\nMissing capabilities → .learnings/FEATURE_REQUESTS.md when you ask for something the skill can't do" }, { "title": "After Each Run", "body": "A SWARM_SUMMARY.md is generated with:\n\nRole success/failure rates\nTotal files generated\nReferences to learnings captured\nRecommendations for next steps" }, { "title": "Promoting Learnings", "body": "Over time, review .learnings/ files:\n\nRecurring error patterns → update orchestrator prompts or add retry logic\nBetter approaches → incorporate into the skill's default behavior\nFeature requests → consider for skill enhancements\n\nThis creates a feedback loop where each swarm run makes the skill smarter." }, { "title": "Example Prompts", "body": "\"Build a Node.js API with Express that serves Moltbook stats from JSON logs\"\n\"Create a React dashboard with dark theme and charts for ClawCredit status\"\n\"Make a CLI tool that checks ClawCredit pre-qualification and notifies via desktop alert\"\n\"Generate a smart contract that holds ClawCredit limits and allows x402 payments\"\n\"Build a hackathon app: a React dashboard that shows user's token balance using Privy auth\" (includes Privy integration out of the box)" }, { "title": "Notes", "body": "The skill makes all decisions autonomously: tech stack, file structure, library choices.\nIf a task fails, the orchestrator will retry once with adjusted instructions.\nYou can monitor progress via the sub-agent logs in .openclaw/agents//sessions/.\nTo stop early, send /stop to the orchestrator's session.\nPrivy Integration: When the prompt mentions blockchain, web3, tokens, NFTs, or Privy, the skill automatically includes Privy authentication and wallet infrastructure. Backend includes /auth/callback with JWKS verification and a simulated fallback; frontend integrates @privy-io/react-auth if React is used. For advanced agentic wallet controls, see the Privy Agentic Wallets skill.\nProject Memory: Each swarm run creates a DECISIONS.md file that documents significant decisions made by the planner and each agent. This serves as long-term knowledge grounding—future developers (or the same human weeks later) can understand why certain choices were made. Agents are prompted to explain their technical decisions (e.g., library selection, architecture patterns, security tradeoffs) as part of their output.\n\nEnjoy your autonomous coding factory 🚀" } ], "body": "Swarm Coding Skill\n\nFully autonomous multi-agent software development. Given a plain-English prompt, the swarm designs, implements, tests, and delivers a complete project end-to-end.\n\nCore capability: Code generation via OpenRouter's qwen3-coder model. The orchestrator drives a Planner to create a manifest, then executes specialized worker roles (BackendDev, FrontendDev, QA, DevOps, etc.) in dependency order. All code is written to files; no interactive sessions.\n\nImportant: This skill generates code for review and deployment by the user. It does not make business decisions or operate autonomously in production. The user remains responsible for security, compliance, and operational decisions.\n\nHow It Works\nOrchestrator (Planner role) analyzes your prompt, decides tech stack and architecture, and creates a swarm.yaml manifest with tasks and dependencies.\nWorker agents (BackendDev, FrontendDev, QA, DevOps) are spawned as sub-sessions. Each has a clear persona and works on its assigned files in a shared workspace.\nCoordination: The orchestrator tracks task completion and dependencies. When a task finishes, it marks it done and starts any unblocked downstream tasks.\nConflict avoidance: Files are partitioned by role (Backend owns server/, Frontend owns client/, etc.). If two roles need the same file, the manifest assigns an owner.\nQuality gates: QA must pass tests before integration; DevOps ensures containerization; no merge without green tests.\nDeliverable: You get a complete project directory with README, tests, Dockerfile, and optionally a GitHub repo or zip.\nUsage\n# In your main OpenClaw session, invoke:\n/trigger swarm-code \"Build a dashboard that shows Moltbook stats and ClawCredit status\"\n\n\nThe skill will:\n\nSpawn the orchestrator in an isolated session\nOrchestrator spawns workers sequentially or in parallel (based on dependencies)\nOutput a final summary and path to the completed project\nRequirements\nNode.js v18+\nEnvironment variables (in .env at workspace root):\nRequired: OPENROUTER_API_KEY — OpenRouter API key with qwen/qwen3-coder access\nOptional: OPENROUTER_MODEL (default: qwen/qwen3-coder), MOCK=1 for dry-run\nInternet access for OpenRouter API (and optionally GitHub/Docker if deployment requested)\n\nImportant: The orchestrator reads .env from the workspace root (parent directory of this skill) and writes project files to swarm-projects/ and logs to .learnings/ in that same workspace root. Run in an isolated workspace to avoid exposing unrelated secrets.\n\nConfiguration\n\nStore your OpenRouter key in .env at the workspace root:\n\nOPENROUTER_API_KEY=sk-or-...\n\n\nOptional overrides:\n\nOPENROUTER_MODEL=qwen/qwen3-coder\nMOCK=1 # dry-run, no API calls\n\n\nThe skill uses qwen/qwen3-coder by default. Ensure your OpenRouter key has that model enabled.\n\nOutput\n\nThe created project lives in swarm-projects// and includes:\n\nREADME.md with run instructions\npackage.json (or equivalent)\nSource code organized by component\ntest/ directory with automated tests\nDockerfile and docker-compose.yml (if applicable)\nCI/ with GitHub Actions workflow (optional)\nDECISIONS.md — Project memory documenting key architectural and technical decisions with rationale\n.learnings/ — Learning logs capturing errors, insights, and feature requests\nERRORS.md — Failures, exceptions, and recovery actions\nLEARNINGS.md — Corrections, better approaches, knowledge gaps\nFEATURE_REQUESTS.md — Requested capabilities that don't exist yet\nSWARM_SUMMARY.md — Execution summary with role performance, statistics, and next steps\nContinuous Improvement\n\nThe swarm skill automatically captures learnings during execution to improve future runs:\n\nWhat Gets Logged\nWorker failures → .learnings/ERRORS.md with context and recovery suggestions\nBetter approaches discovered → .learnings/LEARNINGS.md (e.g., \"Simplified X by using Y\")\nUser corrections → .learnings/LEARNINGS.md when you override a decision\nMissing capabilities → .learnings/FEATURE_REQUESTS.md when you ask for something the skill can't do\nAfter Each Run\n\nA SWARM_SUMMARY.md is generated with:\n\nRole success/failure rates\nTotal files generated\nReferences to learnings captured\nRecommendations for next steps\nPromoting Learnings\n\nOver time, review .learnings/ files:\n\nRecurring error patterns → update orchestrator prompts or add retry logic\nBetter approaches → incorporate into the skill's default behavior\nFeature requests → consider for skill enhancements\n\nThis creates a feedback loop where each swarm run makes the skill smarter.\n\nExample Prompts\n\"Build a Node.js API with Express that serves Moltbook stats from JSON logs\"\n\"Create a React dashboard with dark theme and charts for ClawCredit status\"\n\"Make a CLI tool that checks ClawCredit pre-qualification and notifies via desktop alert\"\n\"Generate a smart contract that holds ClawCredit limits and allows x402 payments\"\n\"Build a hackathon app: a React dashboard that shows user's token balance using Privy auth\" (includes Privy integration out of the box)\nNotes\nThe skill makes all decisions autonomously: tech stack, file structure, library choices.\nIf a task fails, the orchestrator will retry once with adjusted instructions.\nYou can monitor progress via the sub-agent logs in .openclaw/agents//sessions/.\nTo stop early, send /stop to the orchestrator's session.\nPrivy Integration: When the prompt mentions blockchain, web3, tokens, NFTs, or Privy, the skill automatically includes Privy authentication and wallet infrastructure. Backend includes /auth/callback with JWKS verification and a simulated fallback; frontend integrates @privy-io/react-auth if React is used. For advanced agentic wallet controls, see the Privy Agentic Wallets skill.\nProject Memory: Each swarm run creates a DECISIONS.md file that documents significant decisions made by the planner and each agent. This serves as long-term knowledge grounding—future developers (or the same human weeks later) can understand why certain choices were made. Agents are prompted to explain their technical decisions (e.g., library selection, architecture patterns, security tradeoffs) as part of their output.\n\nEnjoy your autonomous coding factory 🚀" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/arunnadarasa/swarm-coding-skill", "publisherUrl": "https://clawhub.ai/arunnadarasa/swarm-coding-skill", "owner": "arunnadarasa", "version": "0.1.2", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/swarm-coding-skill", "downloadUrl": "https://openagent3.xyz/downloads/swarm-coding-skill", "agentUrl": "https://openagent3.xyz/skills/swarm-coding-skill/agent", "manifestUrl": "https://openagent3.xyz/skills/swarm-coding-skill/agent.json", "briefUrl": "https://openagent3.xyz/skills/swarm-coding-skill/agent.md" } }