{
  "schemaVersion": "1.0",
  "item": {
    "slug": "claude-code-teams",
    "name": "Claude Code Agent Teams",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/matthew-a-gordon/claude-code-teams",
    "canonicalUrl": "https://clawhub.ai/matthew-a-gordon/claude-code-teams",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/claude-code-teams",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=claude-code-teams",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "README.md",
      "RESEARCH-SYNTHESIS.md",
      "SKILL.md",
      "package.json",
      "scripts/team-cleanup.sh",
      "scripts/team-monitor.sh"
    ],
    "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": "2026-04-23T16:43:11.935Z",
      "expiresAt": "2026-04-30T16:43:11.935Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard",
        "contentDisposition": "attachment; filename=\"4claw-imageboard-1.0.1.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/claude-code-teams"
    },
    "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/claude-code-teams",
    "agentPageUrl": "https://openagent3.xyz/skills/claude-code-teams/agent",
    "manifestUrl": "https://openagent3.xyz/skills/claude-code-teams/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/claude-code-teams/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": "Claude Code Agent Teams",
        "body": "Orchestrate multiple Claude Code agents working together on complex tasks. Agents coordinate through a shared task list, message each other directly, and work in parallel on independent scopes."
      },
      {
        "title": "When to Use",
        "body": "✅ Use for parallel independent work:\n\nMulti-lens code reviews (security, performance, tests in parallel)\nDebugging with competing hypotheses (test theories simultaneously)\nFull-stack features (frontend/backend/tests by separate specialists)\nArchitecture decisions (adversarial debate produces stronger ADRs)\nCross-domain investigations (bottleneck analysis across systems)\nData-parallel work (inventory classification, bulk refactoring)\n\n❌ Don't use for:\n\nSequential work with tight dependencies\nSame-file edits (conflict risk)\nSimple tasks (coordination overhead > benefit)\nLearning/exploration (single agent with full context is better)\n\nRule of thumb: Teams multiply token cost 3-4x. Use only when parallelization benefit exceeds coordination overhead."
      },
      {
        "title": "1. Enable Feature",
        "body": "Environment variable:\n\nexport CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1\n\nOr settings file (~/.claude/settings.json):\n\n{\n  \"env\": {\n    \"CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS\": \"1\"\n  }\n}"
      },
      {
        "title": "2. Validate Setup",
        "body": "./scripts/validate-setup.sh\n\nChecks for: env var set, claude CLI available, tmux installed (optional)."
      },
      {
        "title": "3. Pick Template",
        "body": "TemplateUse Caseparallel-review.mdMulti-lens code review (security + performance + tests)competing-hypotheses.mdDebug by testing competing theoriesfullstack-feature.mdFrontend/backend/tests coordinationarchitecture-decision.mdADR with adversarial debatebottleneck-analysis.mdCross-domain performance investigationinventory-classification.mdData-parallel classification/refactoring\n\nEach template includes: when to use, team structure, copy-paste spawn prompt, expected outcomes."
      },
      {
        "title": "4. Spawn Team",
        "body": "claude --pty\n# Paste spawn prompt from template\n\nMonitor with ./scripts/team-monitor.sh (separate terminal) or Ctrl+T (in Claude Code)."
      },
      {
        "title": "Core Principles",
        "body": "Clear file boundaries - Each teammate owns different files/directories (prevents conflicts)\nUse delegate mode - Lead coordinates, teammates implement (include in spawn prompt)\nSize tasks appropriately - 5-6 self-contained tasks per teammate\nWait for completion - Don't let lead merge results while teammates are mid-task\nStart with research - Spawn researchers before builders to establish shared understanding"
      },
      {
        "title": "Cost Management",
        "body": "Agent teams multiply token usage 3-4x vs single session. Each teammate has its own context window.\n\nOptimize with:\n\nModel mixing - Opus for lead (strategic), Sonnet for teammates (tactical) = ~40% savings\nMinimize broadcasts - Direct messages only (broadcasts copy to all teammates)\nRight-size teams - 3 teammates usually optimal; 6+ only for highly parallel work\n\nSee references/cost-management.md for detailed strategies."
      },
      {
        "title": "Troubleshooting",
        "body": "Teammates not spawning?\n\nVerify CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 is set\nRestart Claude Code after setting env var\n\nFile conflicts?\n\nDefine explicit file boundaries in spawn prompt (directory-level ownership)\n\nLead implementing instead of coordinating?\n\nAdd \"Use delegate mode: coordinate but don't implement\" to spawn prompt\nOr press Shift+Tab to cycle into delegate mode\n\nSee references/troubleshooting.md for full list."
      },
      {
        "title": "Automation",
        "body": "Monitor progress:\n\n./scripts/team-monitor.sh\n\nCleanup (graceful shutdown):\n\n./scripts/team-cleanup.sh"
      },
      {
        "title": "Display Modes",
        "body": "In-process (default) - All teammates in main terminal\n\nShift+Up/Down - Select teammate\nCtrl+T - Toggle task list\nEscape - Interrupt teammate\n\nSplit-pane (requires tmux or iTerm2) - Each teammate gets own pane\n\nSet teammateMode: \"tmux\" in ~/.claude/settings.json"
      },
      {
        "title": "References",
        "body": "templates/ - 6 copy-paste ready spawn prompts with usage guidance\nscripts/ - Automation helpers (validate, monitor, cleanup)\nreferences/best-practices.md - When to use teams, task sizing, file boundaries\nreferences/cost-management.md - Token optimization strategies\nreferences/troubleshooting.md - Common issues and solutions\nexamples/ - Real-world walkthroughs (PR review, bug hunt, feature build)"
      },
      {
        "title": "Comparison",
        "body": "ApproachBest ForSingle sessionSequential work, explorationAgent teams (this skill)Parallel independent tasks with coordinationOpenClaw subagentsBackground work, long-running tasks"
      },
      {
        "title": "Contributing",
        "body": "Found a useful pattern? Submit PR with:\n\nTemplate or example\nWhen to use + team structure\nSpawn prompt + expected outcomes"
      },
      {
        "title": "Credits",
        "body": "Built with insights from Claude Code documentation, ClaudeFast patterns, and OpenClaw community.\n\nRepository: https://github.com/matthew-a-gordon/claude-code-teams\nLicense: MIT"
      }
    ],
    "body": "Claude Code Agent Teams\n\nOrchestrate multiple Claude Code agents working together on complex tasks. Agents coordinate through a shared task list, message each other directly, and work in parallel on independent scopes.\n\nWhen to Use\n\n✅ Use for parallel independent work:\n\nMulti-lens code reviews (security, performance, tests in parallel)\nDebugging with competing hypotheses (test theories simultaneously)\nFull-stack features (frontend/backend/tests by separate specialists)\nArchitecture decisions (adversarial debate produces stronger ADRs)\nCross-domain investigations (bottleneck analysis across systems)\nData-parallel work (inventory classification, bulk refactoring)\n\n❌ Don't use for:\n\nSequential work with tight dependencies\nSame-file edits (conflict risk)\nSimple tasks (coordination overhead > benefit)\nLearning/exploration (single agent with full context is better)\n\nRule of thumb: Teams multiply token cost 3-4x. Use only when parallelization benefit exceeds coordination overhead.\n\nQuick Start\n1. Enable Feature\n\nEnvironment variable:\n\nexport CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1\n\n\nOr settings file (~/.claude/settings.json):\n\n{\n  \"env\": {\n    \"CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS\": \"1\"\n  }\n}\n\n2. Validate Setup\n./scripts/validate-setup.sh\n\n\nChecks for: env var set, claude CLI available, tmux installed (optional).\n\n3. Pick Template\nTemplate\tUse Case\nparallel-review.md\tMulti-lens code review (security + performance + tests)\ncompeting-hypotheses.md\tDebug by testing competing theories\nfullstack-feature.md\tFrontend/backend/tests coordination\narchitecture-decision.md\tADR with adversarial debate\nbottleneck-analysis.md\tCross-domain performance investigation\ninventory-classification.md\tData-parallel classification/refactoring\n\nEach template includes: when to use, team structure, copy-paste spawn prompt, expected outcomes.\n\n4. Spawn Team\nclaude --pty\n# Paste spawn prompt from template\n\n\nMonitor with ./scripts/team-monitor.sh (separate terminal) or Ctrl+T (in Claude Code).\n\nCore Principles\nClear file boundaries - Each teammate owns different files/directories (prevents conflicts)\nUse delegate mode - Lead coordinates, teammates implement (include in spawn prompt)\nSize tasks appropriately - 5-6 self-contained tasks per teammate\nWait for completion - Don't let lead merge results while teammates are mid-task\nStart with research - Spawn researchers before builders to establish shared understanding\nCost Management\n\nAgent teams multiply token usage 3-4x vs single session. Each teammate has its own context window.\n\nOptimize with:\n\nModel mixing - Opus for lead (strategic), Sonnet for teammates (tactical) = ~40% savings\nMinimize broadcasts - Direct messages only (broadcasts copy to all teammates)\nRight-size teams - 3 teammates usually optimal; 6+ only for highly parallel work\n\nSee references/cost-management.md for detailed strategies.\n\nTroubleshooting\n\nTeammates not spawning?\n\nVerify CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 is set\nRestart Claude Code after setting env var\n\nFile conflicts?\n\nDefine explicit file boundaries in spawn prompt (directory-level ownership)\n\nLead implementing instead of coordinating?\n\nAdd \"Use delegate mode: coordinate but don't implement\" to spawn prompt\nOr press Shift+Tab to cycle into delegate mode\n\nSee references/troubleshooting.md for full list.\n\nAutomation\n\nMonitor progress:\n\n./scripts/team-monitor.sh\n\n\nCleanup (graceful shutdown):\n\n./scripts/team-cleanup.sh\n\nDisplay Modes\n\nIn-process (default) - All teammates in main terminal\n\nShift+Up/Down - Select teammate\nCtrl+T - Toggle task list\nEscape - Interrupt teammate\n\nSplit-pane (requires tmux or iTerm2) - Each teammate gets own pane\n\nSet teammateMode: \"tmux\" in ~/.claude/settings.json\nReferences\ntemplates/ - 6 copy-paste ready spawn prompts with usage guidance\nscripts/ - Automation helpers (validate, monitor, cleanup)\nreferences/best-practices.md - When to use teams, task sizing, file boundaries\nreferences/cost-management.md - Token optimization strategies\nreferences/troubleshooting.md - Common issues and solutions\nexamples/ - Real-world walkthroughs (PR review, bug hunt, feature build)\nComparison\nApproach\tBest For\nSingle session\tSequential work, exploration\nAgent teams (this skill)\tParallel independent tasks with coordination\nOpenClaw subagents\tBackground work, long-running tasks\nContributing\n\nFound a useful pattern? Submit PR with:\n\nTemplate or example\nWhen to use + team structure\nSpawn prompt + expected outcomes\nCredits\n\nBuilt with insights from Claude Code documentation, ClaudeFast patterns, and OpenClaw community.\n\nRepository: https://github.com/matthew-a-gordon/claude-code-teams\nLicense: MIT"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/matthew-a-gordon/claude-code-teams",
    "publisherUrl": "https://clawhub.ai/matthew-a-gordon/claude-code-teams",
    "owner": "matthew-a-gordon",
    "version": "1.0.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/claude-code-teams",
    "downloadUrl": "https://openagent3.xyz/downloads/claude-code-teams",
    "agentUrl": "https://openagent3.xyz/skills/claude-code-teams/agent",
    "manifestUrl": "https://openagent3.xyz/skills/claude-code-teams/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/claude-code-teams/agent.md"
  }
}