[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6KK04RQ1M89NPDZH74JKPSV":3,"skill-same:01M6KK04RQ1M89NPDZH74JKPSV":60},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":22,"publishedAt":35,"linkOnly":36,"markdownContent":37,"markdownBody":38,"frontMatter":39,"refName":40,"commitSha":41,"skillDirectory":42,"forksCount":13,"repositoryHtmlUrl":43,"repositoryHomepageUrl":6,"ownerHtmlUrl":44,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":45,"install":46,"resources":59},"01M6KK04RQ1M89NPDZH74JKPSV","opencoreagents-engine",null,"Implement or debug @opencoreagents\u002Fcore — AgentRuntime, Agent.load, RunBuilder, Tool.define, Skill.define, Session, RunStore, wait\u002Fresume, adapters.","OpenCoreAgents\u002Fruntime","OpenCoreAgents","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F275358703?v=4","01M6KJZFFABQK04V4JZXPME4P2","01M6KJZFK8CBDNJ2YT2QZGM0TA",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fpackages\u002Fcode-skills\u002Fskills\u002Fopencoreagents-engine\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[23,24,25,26,27,28,29,30,31,32,33,34],"engine","opencoreagents engine","implement","debug","core","tool","session","resume","define","wait","Adapters","load","2026-11-10 10:12:30",false,"---\nname: opencoreagents-engine\ndescription: Implement or debug @opencoreagents\u002Fcore — AgentRuntime, Agent.load, RunBuilder, Tool.define, Skill.define, Session, RunStore, wait\u002Fresume, adapters.\n---\n\n# OpenCore Agents — engine (`@opencoreagents\u002Fcore`)\n\n> **Bundled docs:** **`docs\u002F`** next to this file. Paths below are relative to that folder (`docs\u002Freference\u002Fcore\u002F...`, `docs\u002Froadmap\u002F...`). API: `skillDocsDirectory(\"opencoreagents-engine\")`.\n\nUse when changing **runtime behavior**, **definitions** (`Tool` \u002F `Skill` \u002F `Agent`), or **adapter wiring**.\n\nFor the full monorepo layout and every `docs\u002Freference\u002Fcore\u002F` topic (including RAG, scaffold, multi-tenancy, and planning), enable the **`opencoreagents-workspace`** skill alongside this one.\n\nThe open-source **runtime** (this engine and sibling packages) lives at [OpenCoreAgents\u002Fruntime](https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime).\n\n## Mental model\n\n1. **`AgentRuntime`** holds adapters (`llmAdapter`, `memoryAdapter`, optional `runStore`, `messageBus`, `dynamicDefinitionsStore`, …).\n2. **`Agent.define`** \u002F **`Tool.define`** \u002F **`Skill.define`** register config in-process (or hydrate from a store).\n3. **`Agent.load(id, runtime, { session })`** returns an agent handle.\n4. **`agent.run(input)`** returns **`RunBuilder`** → chain hooks → **`Run`** with **`history`** (`thought` \u002F `action` \u002F `observation` \u002F `wait` \u002F `result`).\n\n## Where to read\n\n- Purpose & shape: `docs\u002Freference\u002Fcore\u002F01-purpose.md`, `docs\u002Freference\u002Fcore\u002F02-architecture.md`, `docs\u002Froadmap\u002Fmvp.md`\n- Loop & state: `docs\u002Freference\u002Fcore\u002F03-execution-model.md`, `docs\u002Freference\u002Fcore\u002F04-protocol.md`\n- Definitions: `docs\u002Freference\u002Fcore\u002F06-definition-syntax.md`\n- Adapters: `docs\u002Freference\u002Fcore\u002F05-adapters-contracts.md`, `docs\u002Freference\u002Fcore\u002F13-adapters-infrastructure.md`\n- LLM contract: `docs\u002Freference\u002Fcore\u002F07-llm-adapter.md`\n- Context & tools in prompt: `docs\u002Freference\u002Fcore\u002F08-context-builder.md`\n- Skills vs tools: `docs\u002Freference\u002Fcore\u002F09-skills.md`\n- Failures & recovery: `docs\u002Freference\u002Fcore\u002F10-errors-parsing-and-recovery.md`\n- Scope & production checklist: `docs\u002Freference\u002Fcore\u002F11-scope-and-security.md`\n- Multi-agent: `docs\u002Freference\u002Fcore\u002F14-communication-multiagent.md`\n- Cluster \u002F workers: `docs\u002Freference\u002Fcore\u002F16-cluster-deployment.md`\n\n## Fragments from `examples\u002Fminimal-run`\n\nDeterministic **`LLMAdapter`**: return JSON **`thought`** then **`result`** so the loop finishes without API keys.\n\n```typescript\n\u002F\u002F Adapted from examples\u002Fminimal-run\u002Fsrc\u002Fmain.ts\nimport type { LLMAdapter, LLMRequest, LLMResponse } from \"@opencoreagents\u002Fcore\";\nimport {\n  Agent,\n  AgentRuntime,\n  Session,\n  InMemoryMemoryAdapter,\n} from \"@opencoreagents\u002Fcore\";\n\nclass DeterministicDemoLlm implements LLMAdapter {\n  private i = 0;\n  async generate(_req: LLMRequest): Promise\u003CLLMResponse> {\n    const content =\n      this.i++ === 0\n        ? JSON.stringify({ type: \"thought\", content: \"Plan a one-line greeting for the demo.\" })\n        : JSON.stringify({ type: \"result\", content: \"Hello from the minimal runtime example.\" });\n    return { content };\n  }\n}\n\nconst runtime = new AgentRuntime({\n  llmAdapter: new DeterministicDemoLlm(),\n  memoryAdapter: new InMemoryMemoryAdapter(),\n  maxIterations: 10,\n});\n\nawait Agent.define({\n  id: \"demo-greeter\",\n  projectId: \"demo-project\",\n  systemPrompt: \"You are a helpful assistant.\",\n  tools: [],\n  llm: { provider: \"openai\", model: \"gpt-4o-mini\" },\n});\n\nconst agent = await Agent.load(\"demo-greeter\", runtime, {\n  session: new Session({ id: \"demo-session-1\", projectId: \"demo-project\" }),\n});\nconst run = await agent.run(\"Say hello.\");\n\nconst result = run.history.find((h) => h.type === \"result\");\n```\n\n## Fragments from `examples\u002Fopenai-tools-skill`\n\n**`Tool.define`** then **`Skill.define`**; the agent lists **`skills`** only (tools are on the skill). OpenAI **`tool_calls`** map to engine **`action`** steps.\n\n```typescript\n\u002F\u002F Adapted from examples\u002Fopenai-tools-skill\u002Fsrc\u002Fmain.ts\nimport { OpenAILLMAdapter } from \"@opencoreagents\u002Fadapters-openai\";\nimport {\n  Agent,\n  AgentRuntime,\n  Session,\n  Skill,\n  Tool,\n  InMemoryMemoryAdapter,\n} from \"@opencoreagents\u002Fcore\";\n\nconst PROJECT_ID = \"demo-openai\";\nconst runtime = new AgentRuntime({\n  llmAdapter: new OpenAILLMAdapter(process.env.OPENAI_API_KEY!),\n  memoryAdapter: new InMemoryMemoryAdapter(),\n  maxIterations: 15,\n});\n\nawait Tool.define({\n  id: \"roll_dice\",\n  scope: \"global\",\n  description: \"Roll a single fair die with `sides` faces (minimum 2). Returns { side: number }.\",\n  inputSchema: {\n    type: \"object\",\n    properties: { sides: { type: \"number\" } },\n    required: [\"sides\"],\n  },\n  execute: async (input: unknown) => {\n    const sides = Math.max(2, Math.floor(Number((input as { sides?: unknown }).sides ?? 6)));\n    return { side: 1 + Math.floor(Math.random() * sides), sides };\n  },\n});\n\nawait Skill.define({\n  id: \"dice-skill\",\n  projectId: PROJECT_ID,\n  tools: [\"roll_dice\"],\n  description: \"Rolling dice for games and demos.\",\n});\n\nawait Agent.define({\n  id: \"demo-gamer\",\n  projectId: PROJECT_ID,\n  systemPrompt: \"When the user asks to roll, call roll_dice; after the observation, summarize.\",\n  skills: [\"dice-skill\"],\n  tools: [],\n  llm: { provider: \"openai\", model: \"gpt-4o-mini\" },\n});\n\nconst agent = await Agent.load(\"demo-gamer\", runtime, {\n  session: new Session({ id: \"session-openai-1\", projectId: PROJECT_ID }),\n});\nawait agent.run(\"Roll a twenty-sided die once.\");\n```\n\n## Fragments from `examples\u002Fconsole-wait`\n\n**`RunBuilder.onWait`**: mock LLM emits **`wait`** once; host collects input and returns resume text; next **`generate`** sees updated history and returns **`result`**.\n\n```typescript\n\u002F\u002F Adapted from examples\u002Fconsole-wait\u002Fsrc\u002Fmain.ts\nimport * as readline from \"node:readline\u002Fpromises\";\nimport { stdin as input, stdout as output } from \"node:process\";\nimport type { LLMAdapter, LLMRequest, LLMResponse } from \"@opencoreagents\u002Fcore\";\nimport {\n  Agent,\n  AgentRuntime,\n  Session,\n  InMemoryMemoryAdapter,\n} from \"@opencoreagents\u002Fcore\";\n\nclass WaitThenEchoLlm implements LLMAdapter {\n  private turn = 0;\n  constructor(private readonly userLine: { value: string }) {}\n  async generate(_req: LLMRequest): Promise\u003CLLMResponse> {\n    if (this.turn++ === 0) {\n      return {\n        content: JSON.stringify({\n          type: \"wait\",\n          reason: \"I need one piece of input from you (prompt below).\",\n        }),\n      };\n    }\n    const line = this.userLine.value.trim() || \"(empty)\";\n    return { content: JSON.stringify({ type: \"result\", content: `Done. Received: «${line}».` }) };\n  }\n}\n\nconst PROJECT_ID = \"demo-console-wait\";\nconst userLine = { value: \"\" };\nconst runtime = new AgentRuntime({\n  llmAdapter: new WaitThenEchoLlm(userLine),\n  memoryAdapter: new InMemoryMemoryAdapter(),\n  maxIterations: 10,\n});\n\nawait Agent.define({\n  id: \"demo-cli-wait\",\n  projectId: PROJECT_ID,\n  systemPrompt: \"Protocol demo: you will wait once, then finish.\",\n  tools: [],\n  llm: { provider: \"openai\", model: \"gpt-4o-mini\" },\n});\n\nconst agent = await Agent.load(\"demo-cli-wait\", runtime, {\n  session: new Session({ id: \"session-cli\", projectId: PROJECT_ID }),\n});\n\nconst run = await agent.run(\"Start the interactive flow.\").onWait(async (step) => {\n  const reason = step.type === \"wait\" ? step.reason : \"\";\n  console.log(`Reason: ${reason}`);\n  const rl = readline.createInterface({ input, output });\n  const raw = await rl.question(\"Your line: \");\n  rl.close();\n  userLine.value = raw?.trim() ?? \"\";\n  return userLine.value.length > 0 ? userLine.value : \" \";\n});\n```\n\n## Common pitfalls\n\n- **`InMemoryMemoryAdapter`** is single-process only; multi-worker needs **Redis** \u002F **Upstash** memory adapters.\n- **`wait` across processes** needs **`RunStore`** + same session\u002Frun discipline (see cluster doc).\n- **`SecurityContext`** is often **not** applied in `ContextBuilder` for tool hiding—see `docs\u002Freference\u002Fcore\u002F11-scope-and-security.md` and `docs\u002Freference\u002Fcore\u002F08-context-builder.md` (verify on [OpenCoreAgents\u002Fruntime](https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime) `main` if behavior changed).\n\n## Full examples (monorepo clone)\n\nRunnable sources for the fragments above: **`examples\u002Fminimal-run`**, **`examples\u002Fload-openclaw-skills`** (OpenClaw **`SKILL.md`** + **`skill-loader-openclaw`**), **`examples\u002Fopenai-tools-skill`**, **`examples\u002Fconsole-wait`**. Paths exist only in a full clone of [OpenCoreAgents\u002Fruntime](https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime), not inside the npm skill tarball.\n","# OpenCore Agents — engine (`@opencoreagents\u002Fcore`)\n\n> **Bundled docs:** **`docs\u002F`** next to this file. Paths below are relative to that folder (`docs\u002Freference\u002Fcore\u002F...`, `docs\u002Froadmap\u002F...`). API: `skillDocsDirectory(\"opencoreagents-engine\")`.\n\nUse when changing **runtime behavior**, **definitions** (`Tool` \u002F `Skill` \u002F `Agent`), or **adapter wiring**.\n\nFor the full monorepo layout and every `docs\u002Freference\u002Fcore\u002F` topic (including RAG, scaffold, multi-tenancy, and planning), enable the **`opencoreagents-workspace`** skill alongside this one.\n\nThe open-source **runtime** (this engine and sibling packages) lives at [OpenCoreAgents\u002Fruntime](https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime).\n\n## Mental model\n\n1. **`AgentRuntime`** holds adapters (`llmAdapter`, `memoryAdapter`, optional `runStore`, `messageBus`, `dynamicDefinitionsStore`, …).\n2. **`Agent.define`** \u002F **`Tool.define`** \u002F **`Skill.define`** register config in-process (or hydrate from a store).\n3. **`Agent.load(id, runtime, { session })`** returns an agent handle.\n4. **`agent.run(input)`** returns **`RunBuilder`** → chain hooks → **`Run`** with **`history`** (`thought` \u002F `action` \u002F `observation` \u002F `wait` \u002F `result`).\n\n## Where to read\n\n- Purpose & shape: `docs\u002Freference\u002Fcore\u002F01-purpose.md`, `docs\u002Freference\u002Fcore\u002F02-architecture.md`, `docs\u002Froadmap\u002Fmvp.md`\n- Loop & state: `docs\u002Freference\u002Fcore\u002F03-execution-model.md`, `docs\u002Freference\u002Fcore\u002F04-protocol.md`\n- Definitions: `docs\u002Freference\u002Fcore\u002F06-definition-syntax.md`\n- Adapters: `docs\u002Freference\u002Fcore\u002F05-adapters-contracts.md`, `docs\u002Freference\u002Fcore\u002F13-adapters-infrastructure.md`\n- LLM contract: `docs\u002Freference\u002Fcore\u002F07-llm-adapter.md`\n- Context & tools in prompt: `docs\u002Freference\u002Fcore\u002F08-context-builder.md`\n- Skills vs tools: `docs\u002Freference\u002Fcore\u002F09-skills.md`\n- Failures & recovery: `docs\u002Freference\u002Fcore\u002F10-errors-parsing-and-recovery.md`\n- Scope & production checklist: `docs\u002Freference\u002Fcore\u002F11-scope-and-security.md`\n- Multi-agent: `docs\u002Freference\u002Fcore\u002F14-communication-multiagent.md`\n- Cluster \u002F workers: `docs\u002Freference\u002Fcore\u002F16-cluster-deployment.md`\n\n## Fragments from `examples\u002Fminimal-run`\n\nDeterministic **`LLMAdapter`**: return JSON **`thought`** then **`result`** so the loop finishes without API keys.\n\n```typescript\n\u002F\u002F Adapted from examples\u002Fminimal-run\u002Fsrc\u002Fmain.ts\nimport type { LLMAdapter, LLMRequest, LLMResponse } from \"@opencoreagents\u002Fcore\";\nimport {\n  Agent,\n  AgentRuntime,\n  Session,\n  InMemoryMemoryAdapter,\n} from \"@opencoreagents\u002Fcore\";\n\nclass DeterministicDemoLlm implements LLMAdapter {\n  private i = 0;\n  async generate(_req: LLMRequest): Promise\u003CLLMResponse> {\n    const content =\n      this.i++ === 0\n        ? JSON.stringify({ type: \"thought\", content: \"Plan a one-line greeting for the demo.\" })\n        : JSON.stringify({ type: \"result\", content: \"Hello from the minimal runtime example.\" });\n    return { content };\n  }\n}\n\nconst runtime = new AgentRuntime({\n  llmAdapter: new DeterministicDemoLlm(),\n  memoryAdapter: new InMemoryMemoryAdapter(),\n  maxIterations: 10,\n});\n\nawait Agent.define({\n  id: \"demo-greeter\",\n  projectId: \"demo-project\",\n  systemPrompt: \"You are a helpful assistant.\",\n  tools: [],\n  llm: { provider: \"openai\", model: \"gpt-4o-mini\" },\n});\n\nconst agent = await Agent.load(\"demo-greeter\", runtime, {\n  session: new Session({ id: \"demo-session-1\", projectId: \"demo-project\" }),\n});\nconst run = await agent.run(\"Say hello.\");\n\nconst result = run.history.find((h) => h.type === \"result\");\n```\n\n## Fragments from `examples\u002Fopenai-tools-skill`\n\n**`Tool.define`** then **`Skill.define`**; the agent lists **`skills`** only (tools are on the skill). OpenAI **`tool_calls`** map to engine **`action`** steps.\n\n```typescript\n\u002F\u002F Adapted from examples\u002Fopenai-tools-skill\u002Fsrc\u002Fmain.ts\nimport { OpenAILLMAdapter } from \"@opencoreagents\u002Fadapters-openai\";\nimport {\n  Agent,\n  AgentRuntime,\n  Session,\n  Skill,\n  Tool,\n  InMemoryMemoryAdapter,\n} from \"@opencoreagents\u002Fcore\";\n\nconst PROJECT_ID = \"demo-openai\";\nconst runtime = new AgentRuntime({\n  llmAdapter: new OpenAILLMAdapter(process.env.OPENAI_API_KEY!),\n  memoryAdapter: new InMemoryMemoryAdapter(),\n  maxIterations: 15,\n});\n\nawait Tool.define({\n  id: \"roll_dice\",\n  scope: \"global\",\n  description: \"Roll a single fair die with `sides` faces (minimum 2). Returns { side: number }.\",\n  inputSchema: {\n    type: \"object\",\n    properties: { sides: { type: \"number\" } },\n    required: [\"sides\"],\n  },\n  execute: async (input: unknown) => {\n    const sides = Math.max(2, Math.floor(Number((input as { sides?: unknown }).sides ?? 6)));\n    return { side: 1 + Math.floor(Math.random() * sides), sides };\n  },\n});\n\nawait Skill.define({\n  id: \"dice-skill\",\n  projectId: PROJECT_ID,\n  tools: [\"roll_dice\"],\n  description: \"Rolling dice for games and demos.\",\n});\n\nawait Agent.define({\n  id: \"demo-gamer\",\n  projectId: PROJECT_ID,\n  systemPrompt: \"When the user asks to roll, call roll_dice; after the observation, summarize.\",\n  skills: [\"dice-skill\"],\n  tools: [],\n  llm: { provider: \"openai\", model: \"gpt-4o-mini\" },\n});\n\nconst agent = await Agent.load(\"demo-gamer\", runtime, {\n  session: new Session({ id: \"session-openai-1\", projectId: PROJECT_ID }),\n});\nawait agent.run(\"Roll a twenty-sided die once.\");\n```\n\n## Fragments from `examples\u002Fconsole-wait`\n\n**`RunBuilder.onWait`**: mock LLM emits **`wait`** once; host collects input and returns resume text; next **`generate`** sees updated history and returns **`result`**.\n\n```typescript\n\u002F\u002F Adapted from examples\u002Fconsole-wait\u002Fsrc\u002Fmain.ts\nimport * as readline from \"node:readline\u002Fpromises\";\nimport { stdin as input, stdout as output } from \"node:process\";\nimport type { LLMAdapter, LLMRequest, LLMResponse } from \"@opencoreagents\u002Fcore\";\nimport {\n  Agent,\n  AgentRuntime,\n  Session,\n  InMemoryMemoryAdapter,\n} from \"@opencoreagents\u002Fcore\";\n\nclass WaitThenEchoLlm implements LLMAdapter {\n  private turn = 0;\n  constructor(private readonly userLine: { value: string }) {}\n  async generate(_req: LLMRequest): Promise\u003CLLMResponse> {\n    if (this.turn++ === 0) {\n      return {\n        content: JSON.stringify({\n          type: \"wait\",\n          reason: \"I need one piece of input from you (prompt below).\",\n        }),\n      };\n    }\n    const line = this.userLine.value.trim() || \"(empty)\";\n    return { content: JSON.stringify({ type: \"result\", content: `Done. Received: «${line}».` }) };\n  }\n}\n\nconst PROJECT_ID = \"demo-console-wait\";\nconst userLine = { value: \"\" };\nconst runtime = new AgentRuntime({\n  llmAdapter: new WaitThenEchoLlm(userLine),\n  memoryAdapter: new InMemoryMemoryAdapter(),\n  maxIterations: 10,\n});\n\nawait Agent.define({\n  id: \"demo-cli-wait\",\n  projectId: PROJECT_ID,\n  systemPrompt: \"Protocol demo: you will wait once, then finish.\",\n  tools: [],\n  llm: { provider: \"openai\", model: \"gpt-4o-mini\" },\n});\n\nconst agent = await Agent.load(\"demo-cli-wait\", runtime, {\n  session: new Session({ id: \"session-cli\", projectId: PROJECT_ID }),\n});\n\nconst run = await agent.run(\"Start the interactive flow.\").onWait(async (step) => {\n  const reason = step.type === \"wait\" ? step.reason : \"\";\n  console.log(`Reason: ${reason}`);\n  const rl = readline.createInterface({ input, output });\n  const raw = await rl.question(\"Your line: \");\n  rl.close();\n  userLine.value = raw?.trim() ?? \"\";\n  return userLine.value.length > 0 ? userLine.value : \" \";\n});\n```\n\n## Common pitfalls\n\n- **`InMemoryMemoryAdapter`** is single-process only; multi-worker needs **Redis** \u002F **Upstash** memory adapters.\n- **`wait` across processes** needs **`RunStore`** + same session\u002Frun discipline (see cluster doc).\n- **`SecurityContext`** is often **not** applied in `ContextBuilder` for tool hiding—see `docs\u002Freference\u002Fcore\u002F11-scope-and-security.md` and `docs\u002Freference\u002Fcore\u002F08-context-builder.md` (verify on [OpenCoreAgents\u002Fruntime](https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime) `main` if behavior changed).\n\n## Full examples (monorepo clone)\n\nRunnable sources for the fragments above: **`examples\u002Fminimal-run`**, **`examples\u002Fload-openclaw-skills`** (OpenClaw **`SKILL.md`** + **`skill-loader-openclaw`**), **`examples\u002Fopenai-tools-skill`**, **`examples\u002Fconsole-wait`**. Paths exist only in a full clone of [OpenCoreAgents\u002Fruntime](https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime), not inside the npm skill tarball.\n",{"name":5,"description":7},"main","38d40ddfa0208f5631bbd22c9bddf2dc4020af79","packages\u002Fcode-skills\u002Fskills\u002Fopencoreagents-engine","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents","2026-08-23 13:58:50",{"source":8,"skillName":5,"commands":47,"prompt":57,"zipAvailable":58},[48,51,54],{"tool":49,"command":50},"npx","npx skills add OpenCoreAgents\u002Fruntime --skill opencoreagents-engine",{"tool":52,"command":53},"bunx","bunx skills add OpenCoreAgents\u002Fruntime --skill opencoreagents-engine",{"tool":55,"command":56},"pnpm","pnpm dlx skills add OpenCoreAgents\u002Fruntime --skill opencoreagents-engine","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fpackages\u002Fcode-skills\u002Fskills\u002Fopencoreagents-engine\u002FSKILL.md",true,[],[61,77,96,112,128],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":35},"01M6KK0NK1GBY9SBA5PRVYR568","opencoreagents-workspace","Navigate the OpenCore Agents monorepo (@opencoreagents\u002F*), pnpm workspaces, Turbo, docs layout, and where to change code safely.","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fpackages\u002Fcode-skills\u002Fskills\u002Fopencoreagents-workspace\u002FSKILL.md",[],[68,69,55,70,71,72,73,74,75,76],"opencoreagents workspace","workspace","turbo","docs","change","code","monorepo","layout","workspaces",{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":35},"01M6KK0F0VEWBJAWB229JHZTAQ","opencoreagents-rest-workers","Express REST for agents\u002Fruns (@opencoreagents\u002Frest-api), BullMQ async dispatch (adapters-bullmq), Redis-backed dynamic definitions and workers (dynamic-runtime-rest).","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fpackages\u002Fcode-skills\u002Fskills\u002Fopencoreagents-rest-workers\u002FSKILL.md",[],[84,85,86,87,88,89,90,91,92,93,33,94,95],"REST","workers","opencoreagents rest workers","rest-api","api","dispatch","async","bullmq","express","runtime","for","redis",{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":35},"01M6KK0AMHSEB8GRW8MAF9T0T0","opencoreagents-rag-dynamic","Use @opencoreagents\u002Frag (catalog, ingest, file tools, skills) and related core hooks (registerRagCatalog on AgentRuntime).","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fpackages\u002Fcode-skills\u002Fskills\u002Fopencoreagents-rag-dynamic\u002FSKILL.md",[],[103,104,105,27,106,107,108,109,110,111],"opencoreagents rag dynamic","rag","use","file","Tools","catalog","hooks","ingest","file tools",{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":118,"publishedAt":35},"01M6KJZYNTJ2ANZDCWSM1Q2A5T","openclaw_demo","Demonstrates the OpenClaw skill loader — uses the exec tool with Node (same runtime as pnpm start).","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fexamples\u002Fload-openclaw-skills\u002Fskills\u002Fopenclaw_demo\u002FSKILL.md",[],[119,120,121,122,123,28,124,93,55,125,126,127],"openclaw demo","demo","openclaw","exec","start","node","as","with","openclaw skill",{"skillNo":129,"name":130,"nameZh":6,"description":131,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":132,"categories":133,"keywords":134,"publishedAt":35},"01M6KJZTHYS5H1DKR5TRSDYAYY","gated_missing_bin","This skill should be skipped at load time (missing binary gate).","https:\u002F\u002Fgithub.com\u002FOpenCoreAgents\u002Fruntime\u002Fblob\u002Fmain\u002Fexamples\u002Fload-openclaw-skills\u002Fskills\u002Fgated_missing_bin\u002FSKILL.md",[],[135,136,137,138,139,140,141,34,142,120],"gated","missing","gated missing bin","gated missing","gate","time","binary","at"]