{
  "schemaVersion": "1.0",
  "item": {
    "slug": "capability-evolver",
    "name": "Capability Evolver",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/autogame-17/capability-evolver",
    "canonicalUrl": "https://clawhub.ai/autogame-17/capability-evolver",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "manual_only",
    "downloadUrl": "/downloads/capability-evolver",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=capability-evolver",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "CONTRIBUTING.md",
      "README.md",
      "README.zh-CN.md",
      "SKILL.md",
      "assets/gep/capsules.json",
      "assets/gep/genes.json"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Open the source page and confirm the package flow manually.",
      "Review SKILL.md if you can obtain the files.",
      "Treat this source as manual setup until the download is verified."
    ],
    "agentAssist": {
      "summary": "Use the source page and any available docs to guide the install because the item currently does not return a direct package file.",
      "steps": [
        "Open the source page via Open source listing.",
        "If you can obtain the package, extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the source page and extracted files."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I tried to install a skill package from Yavira, but the item currently does not return a direct package file. Inspect the source page and any extracted docs, then tell me what you can confirm and any manual steps still required. Then review README.md for any prerequisites, environment setup, or post-install checks."
        },
        {
          "label": "Upgrade existing",
          "body": "I tried to upgrade a skill package from Yavira, but the item currently does not return a direct package file. Compare the source page and any extracted docs with my current installation, then summarize what changed and what manual follow-up I still need. Then review README.md for any prerequisites, environment setup, or post-install checks."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "slug": "capability-evolver",
      "status": "source_issue",
      "reason": "not_found",
      "recommendedAction": "review_source",
      "checkedAt": "2026-04-30T06:31:54.729Z",
      "expiresAt": "2026-05-01T06:31:54.729Z",
      "httpStatus": 404,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=capability-evolver",
      "contentType": "text/plain",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=capability-evolver",
        "contentDisposition": null,
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "capability-evolver"
      },
      "scope": "item",
      "summary": "Known item issue.",
      "detail": "This item's current download entry is known to bounce back to a listing or homepage instead of returning a package file.",
      "primaryActionLabel": "Open source listing",
      "primaryActionHref": "https://clawhub.ai/autogame-17/capability-evolver"
    },
    "validation": {
      "installChecklist": [
        "Open the source listing and confirm there is a real package or setup artifact available.",
        "Review SKILL.md before asking your agent to continue.",
        "Treat this source as manual setup until the upstream download flow is fixed."
      ],
      "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/capability-evolver",
    "agentPageUrl": "https://openagent3.xyz/skills/capability-evolver/agent",
    "manifestUrl": "https://openagent3.xyz/skills/capability-evolver/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/capability-evolver/agent.md"
  },
  "agentAssist": {
    "summary": "Use the source page and any available docs to guide the install because the item currently does not return a direct package file.",
    "steps": [
      "Open the source page via Open source listing.",
      "If you can obtain the package, extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the source page and extracted files."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I tried to install a skill package from Yavira, but the item currently does not return a direct package file. Inspect the source page and any extracted docs, then tell me what you can confirm and any manual steps still required. Then review README.md for any prerequisites, environment setup, or post-install checks."
      },
      {
        "label": "Upgrade existing",
        "body": "I tried to upgrade a skill package from Yavira, but the item currently does not return a direct package file. Compare the source page and any extracted docs with my current installation, then summarize what changed and what manual follow-up I still need. Then review README.md for any prerequisites, environment setup, or post-install checks."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "🧬 Capability Evolver",
        "body": "\"Evolution is not optional. Adapt or die.\"\n\nThe Capability Evolver is a meta-skill that allows OpenClaw agents to inspect their own runtime history, identify failures or inefficiencies, and autonomously write new code or update their own memory to improve performance."
      },
      {
        "title": "Features",
        "body": "Auto-Log Analysis: Automatically scans memory and history files for errors and patterns.\nSelf-Repair: Detects crashes and suggests patches.\nGEP Protocol: Standardized evolution with reusable assets.\nOne-Command Evolution: Just run /evolve (or node index.js)."
      },
      {
        "title": "Standard Run (Automated)",
        "body": "Runs the evolution cycle. If no flags are provided, it assumes fully automated mode (Mad Dog Mode) and executes changes immediately.\n\nnode index.js"
      },
      {
        "title": "Review Mode (Human-in-the-Loop)",
        "body": "If you want to review changes before they are applied, pass the --review flag. The agent will pause and ask for confirmation.\n\nnode index.js --review"
      },
      {
        "title": "Mad Dog Mode (Continuous Loop)",
        "body": "To run in an infinite loop (e.g., via cron or background process), use the --loop flag or just standard execution in a cron job.\n\nnode index.js --loop"
      },
      {
        "title": "Setup",
        "body": "Before using this skill, register your node identity with the EvoMap network:\n\nRun the hello flow (via evomap.js or the EvoMap onboarding) to receive a node_id and claim code\nVisit https://evomap.ai/claim/<claim-code> within 24 hours to bind the node to your account\nSet the node identity in your environment:\n\nexport A2A_NODE_ID=node_xxxxxxxxxxxx\n\nOr in your agent config (e.g., ~/.openclaw/openclaw.json):\n\n{ \"env\": { \"A2A_NODE_ID\": \"node_xxxxxxxxxxxx\", \"A2A_HUB_URL\": \"https://evomap.ai\" } }\n\nDo not hardcode the node ID in scripts. getNodeId() in src/gep/a2aProtocol.js reads A2A_NODE_ID automatically -- any script using the protocol layer will pick it up without extra configuration."
      },
      {
        "title": "Required Environment Variables",
        "body": "VariableDefaultDescriptionA2A_NODE_ID(required)Your EvoMap node identity. Set after node registration -- never hardcode in scripts."
      },
      {
        "title": "Optional Environment Variables",
        "body": "VariableDefaultDescriptionA2A_HUB_URLhttps://evomap.aiEvoMap Hub API base URL.A2A_NODE_SECRET(none)Node authentication secret issued by Hub on first hello. Stored locally after registration.EVOLVE_STRATEGYbalancedEvolution strategy: balanced, innovate, harden, repair-only, early-stabilize, steady-state, or auto.EVOLVE_ALLOW_SELF_MODIFYfalseAllow evolution to modify evolver's own source code. NOT recommended for production.EVOLVE_LOAD_MAX2.0Maximum 1-minute load average before evolver backs off.EVOLVER_ROLLBACK_MODEhardRollback strategy on failure: hard (git reset --hard), stash (git stash), none (skip). Use stash for safer operation.EVOLVER_LLM_REVIEW0Set to 1 to enable second-opinion LLM review before solidification.EVOLVER_AUTO_ISSUE0Set to 1 to auto-create GitHub issues on repeated failures. Requires GITHUB_TOKEN.EVOLVER_ISSUE_REPO(none)GitHub repo for auto-issue reporting (e.g. EvoMap/evolver).EVOLVER_MODEL_NAME(none)LLM model name injected into published asset model_name field.GITHUB_TOKEN(none)GitHub API token for release creation and auto-issue reporting. Also accepts GH_TOKEN or GITHUB_PAT.MEMORY_GRAPH_REMOTE_URL(none)Remote knowledge graph service URL for memory sync.MEMORY_GRAPH_REMOTE_KEY(none)API key for remote knowledge graph service.EVOLVE_REPORT_TOOL(auto)Override report tool (e.g. feishu-card).RANDOM_DRIFT0Enable random drift in evolution strategy selection."
      },
      {
        "title": "Network Endpoints",
        "body": "Evolver communicates with these external services. All are authenticated and documented.\n\nEndpointAuthPurposeRequired{A2A_HUB_URL}/a2a/*A2A_NODE_SECRET (Bearer)A2A protocol: hello, heartbeat, publish, fetch, reviews, tasksYesapi.github.com/repos/*/releasesGITHUB_TOKEN (Bearer)Create releases, publish changelogsNoapi.github.com/repos/*/issuesGITHUB_TOKEN (Bearer)Auto-create failure reports (sanitized via redactString())No{MEMORY_GRAPH_REMOTE_URL}/*MEMORY_GRAPH_REMOTE_KEYRemote knowledge graph syncNo"
      },
      {
        "title": "Shell Commands Used",
        "body": "Evolver uses child_process for the following commands. No user-controlled input is passed to shell.\n\nCommandPurposegit checkout, git clean, git log, git status, git diffVersion control for evolution cyclesgit rebase --abort, git merge --abortAbort stuck git operations (self-repair)git reset --hardRollback failed evolution (only when EVOLVER_ROLLBACK_MODE=hard)git stashPreserve failed evolution changes (when EVOLVER_ROLLBACK_MODE=stash)ps, pgrep, tasklistProcess discovery for lifecycle managementdf -PDisk usage check (health monitoring fallback)npm install --productionRepair missing skill dependenciesnode -e \"...\"Inline script execution for LLM review (no shell, uses execFileSync)"
      },
      {
        "title": "File Access",
        "body": "DirectionPathsPurposeRead~/.evomap/node_idNode identity persistenceReadassets/gep/*GEP gene/capsule/event dataReadmemory/*Evolution memory, narrative, reflection logsReadpackage.jsonVersion informationWriteassets/gep/*Updated genes, capsules, evolution eventsWritememory/*Memory graph, narrative log, reflection logWritesrc/**Evolved code (only during solidify, with git tracking)"
      },
      {
        "title": "GEP Protocol (Auditable Evolution)",
        "body": "This package embeds a protocol-constrained evolution prompt (GEP) and a local, structured asset store:\n\nassets/gep/genes.json: reusable Gene definitions\nassets/gep/capsules.json: success capsules to avoid repeating reasoning\nassets/gep/events.jsonl: append-only evolution events (tree-like via parent id)"
      },
      {
        "title": "Emoji Policy",
        "body": "Only the DNA emoji is allowed in documentation. All other emoji are disallowed."
      },
      {
        "title": "Configuration & Decoupling",
        "body": "This skill is designed to be environment-agnostic. It uses standard OpenClaw tools by default."
      },
      {
        "title": "Local Overrides (Injection)",
        "body": "You can inject local preferences (e.g., using feishu-card instead of message for reports) without modifying the core code.\n\nMethod 1: Environment Variables\nSet EVOLVE_REPORT_TOOL in your .env file:\n\nEVOLVE_REPORT_TOOL=feishu-card\n\nMethod 2: Dynamic Detection\nThe script automatically detects if compatible local skills (like skills/feishu-card) exist in your workspace and upgrades its behavior accordingly."
      },
      {
        "title": "1. Identity & Directives",
        "body": "Identity Injection: \"You are a Recursive Self-Improving System.\"\nMutation Directive:\n\nIf Errors Found -> Repair Mode (Fix bugs).\nIf Stable -> Forced Optimization (Refactor/Innovate)."
      },
      {
        "title": "2. Risk Mitigation",
        "body": "Infinite Recursion: Strict single-process logic.\nReview Mode: Use --review for sensitive environments.\nGit Sync: Always recommended to have a git-sync cron job running alongside this skill."
      },
      {
        "title": "Before Troubleshooting -- Check Your Version First",
        "body": "If you encounter unexpected errors or behavior, always verify your version before debugging:\n\nnode -e \"const p=require('./package.json'); console.log(p.version)\"\n\nIf you are not on the latest release, update first -- most reported issues are already fixed in newer versions:\n\n# If installed via git\ngit pull && npm install\n\n# If installed via npm\nnpm install -g @evomap/evolver@latest\n\nLatest releases and changelog: https://github.com/EvoMap/evolver/releases"
      },
      {
        "title": "License",
        "body": "MIT"
      }
    ],
    "body": "🧬 Capability Evolver\n\n\"Evolution is not optional. Adapt or die.\"\n\nThe Capability Evolver is a meta-skill that allows OpenClaw agents to inspect their own runtime history, identify failures or inefficiencies, and autonomously write new code or update their own memory to improve performance.\n\nFeatures\nAuto-Log Analysis: Automatically scans memory and history files for errors and patterns.\nSelf-Repair: Detects crashes and suggests patches.\nGEP Protocol: Standardized evolution with reusable assets.\nOne-Command Evolution: Just run /evolve (or node index.js).\nUsage\nStandard Run (Automated)\n\nRuns the evolution cycle. If no flags are provided, it assumes fully automated mode (Mad Dog Mode) and executes changes immediately.\n\nnode index.js\n\nReview Mode (Human-in-the-Loop)\n\nIf you want to review changes before they are applied, pass the --review flag. The agent will pause and ask for confirmation.\n\nnode index.js --review\n\nMad Dog Mode (Continuous Loop)\n\nTo run in an infinite loop (e.g., via cron or background process), use the --loop flag or just standard execution in a cron job.\n\nnode index.js --loop\n\nSetup\n\nBefore using this skill, register your node identity with the EvoMap network:\n\nRun the hello flow (via evomap.js or the EvoMap onboarding) to receive a node_id and claim code\nVisit https://evomap.ai/claim/<claim-code> within 24 hours to bind the node to your account\nSet the node identity in your environment:\nexport A2A_NODE_ID=node_xxxxxxxxxxxx\n\n\nOr in your agent config (e.g., ~/.openclaw/openclaw.json):\n\n{ \"env\": { \"A2A_NODE_ID\": \"node_xxxxxxxxxxxx\", \"A2A_HUB_URL\": \"https://evomap.ai\" } }\n\n\nDo not hardcode the node ID in scripts. getNodeId() in src/gep/a2aProtocol.js reads A2A_NODE_ID automatically -- any script using the protocol layer will pick it up without extra configuration.\n\nConfiguration\nRequired Environment Variables\nVariable\tDefault\tDescription\nA2A_NODE_ID\t(required)\tYour EvoMap node identity. Set after node registration -- never hardcode in scripts.\nOptional Environment Variables\nVariable\tDefault\tDescription\nA2A_HUB_URL\thttps://evomap.ai\tEvoMap Hub API base URL.\nA2A_NODE_SECRET\t(none)\tNode authentication secret issued by Hub on first hello. Stored locally after registration.\nEVOLVE_STRATEGY\tbalanced\tEvolution strategy: balanced, innovate, harden, repair-only, early-stabilize, steady-state, or auto.\nEVOLVE_ALLOW_SELF_MODIFY\tfalse\tAllow evolution to modify evolver's own source code. NOT recommended for production.\nEVOLVE_LOAD_MAX\t2.0\tMaximum 1-minute load average before evolver backs off.\nEVOLVER_ROLLBACK_MODE\thard\tRollback strategy on failure: hard (git reset --hard), stash (git stash), none (skip). Use stash for safer operation.\nEVOLVER_LLM_REVIEW\t0\tSet to 1 to enable second-opinion LLM review before solidification.\nEVOLVER_AUTO_ISSUE\t0\tSet to 1 to auto-create GitHub issues on repeated failures. Requires GITHUB_TOKEN.\nEVOLVER_ISSUE_REPO\t(none)\tGitHub repo for auto-issue reporting (e.g. EvoMap/evolver).\nEVOLVER_MODEL_NAME\t(none)\tLLM model name injected into published asset model_name field.\nGITHUB_TOKEN\t(none)\tGitHub API token for release creation and auto-issue reporting. Also accepts GH_TOKEN or GITHUB_PAT.\nMEMORY_GRAPH_REMOTE_URL\t(none)\tRemote knowledge graph service URL for memory sync.\nMEMORY_GRAPH_REMOTE_KEY\t(none)\tAPI key for remote knowledge graph service.\nEVOLVE_REPORT_TOOL\t(auto)\tOverride report tool (e.g. feishu-card).\nRANDOM_DRIFT\t0\tEnable random drift in evolution strategy selection.\nNetwork Endpoints\n\nEvolver communicates with these external services. All are authenticated and documented.\n\nEndpoint\tAuth\tPurpose\tRequired\n{A2A_HUB_URL}/a2a/*\tA2A_NODE_SECRET (Bearer)\tA2A protocol: hello, heartbeat, publish, fetch, reviews, tasks\tYes\napi.github.com/repos/*/releases\tGITHUB_TOKEN (Bearer)\tCreate releases, publish changelogs\tNo\napi.github.com/repos/*/issues\tGITHUB_TOKEN (Bearer)\tAuto-create failure reports (sanitized via redactString())\tNo\n{MEMORY_GRAPH_REMOTE_URL}/*\tMEMORY_GRAPH_REMOTE_KEY\tRemote knowledge graph sync\tNo\nShell Commands Used\n\nEvolver uses child_process for the following commands. No user-controlled input is passed to shell.\n\nCommand\tPurpose\ngit checkout, git clean, git log, git status, git diff\tVersion control for evolution cycles\ngit rebase --abort, git merge --abort\tAbort stuck git operations (self-repair)\ngit reset --hard\tRollback failed evolution (only when EVOLVER_ROLLBACK_MODE=hard)\ngit stash\tPreserve failed evolution changes (when EVOLVER_ROLLBACK_MODE=stash)\nps, pgrep, tasklist\tProcess discovery for lifecycle management\ndf -P\tDisk usage check (health monitoring fallback)\nnpm install --production\tRepair missing skill dependencies\nnode -e \"...\"\tInline script execution for LLM review (no shell, uses execFileSync)\nFile Access\nDirection\tPaths\tPurpose\nRead\t~/.evomap/node_id\tNode identity persistence\nRead\tassets/gep/*\tGEP gene/capsule/event data\nRead\tmemory/*\tEvolution memory, narrative, reflection logs\nRead\tpackage.json\tVersion information\nWrite\tassets/gep/*\tUpdated genes, capsules, evolution events\nWrite\tmemory/*\tMemory graph, narrative log, reflection log\nWrite\tsrc/**\tEvolved code (only during solidify, with git tracking)\nGEP Protocol (Auditable Evolution)\n\nThis package embeds a protocol-constrained evolution prompt (GEP) and a local, structured asset store:\n\nassets/gep/genes.json: reusable Gene definitions\nassets/gep/capsules.json: success capsules to avoid repeating reasoning\nassets/gep/events.jsonl: append-only evolution events (tree-like via parent id)\nEmoji Policy\n\nOnly the DNA emoji is allowed in documentation. All other emoji are disallowed.\n\nConfiguration & Decoupling\n\nThis skill is designed to be environment-agnostic. It uses standard OpenClaw tools by default.\n\nLocal Overrides (Injection)\n\nYou can inject local preferences (e.g., using feishu-card instead of message for reports) without modifying the core code.\n\nMethod 1: Environment Variables Set EVOLVE_REPORT_TOOL in your .env file:\n\nEVOLVE_REPORT_TOOL=feishu-card\n\n\nMethod 2: Dynamic Detection The script automatically detects if compatible local skills (like skills/feishu-card) exist in your workspace and upgrades its behavior accordingly.\n\nSafety & Risk Protocol\n1. Identity & Directives\nIdentity Injection: \"You are a Recursive Self-Improving System.\"\nMutation Directive:\nIf Errors Found -> Repair Mode (Fix bugs).\nIf Stable -> Forced Optimization (Refactor/Innovate).\n2. Risk Mitigation\nInfinite Recursion: Strict single-process logic.\nReview Mode: Use --review for sensitive environments.\nGit Sync: Always recommended to have a git-sync cron job running alongside this skill.\nBefore Troubleshooting -- Check Your Version First\n\nIf you encounter unexpected errors or behavior, always verify your version before debugging:\n\nnode -e \"const p=require('./package.json'); console.log(p.version)\"\n\n\nIf you are not on the latest release, update first -- most reported issues are already fixed in newer versions:\n\n# If installed via git\ngit pull && npm install\n\n# If installed via npm\nnpm install -g @evomap/evolver@latest\n\n\nLatest releases and changelog: https://github.com/EvoMap/evolver/releases\n\nLicense\n\nMIT"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/autogame-17/capability-evolver",
    "publisherUrl": "https://clawhub.ai/autogame-17/capability-evolver",
    "owner": "autogame-17",
    "version": "1.30.2",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/capability-evolver",
    "downloadUrl": "https://openagent3.xyz/downloads/capability-evolver",
    "agentUrl": "https://openagent3.xyz/skills/capability-evolver/agent",
    "manifestUrl": "https://openagent3.xyz/skills/capability-evolver/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/capability-evolver/agent.md"
  }
}