[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6ESF42K12E95EE3F6SJKHTJ":3,"skill-same:01M6ESF42K12E95EE3F6SJKHTJ":66},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":32,"linkOnly":33,"markdownContent":34,"markdownBody":35,"frontMatter":36,"refName":37,"commitSha":38,"skillDirectory":39,"forksCount":40,"repositoryHtmlUrl":41,"repositoryHomepageUrl":42,"ownerHtmlUrl":43,"ownerBlogUrl":6,"ownerDescription":44,"updatedAt":45,"install":46,"resources":59},"01M6ESF42K12E95EE3F6SJKHTJ","parallel-agents",null,"Dispatch multiple agents to work on independent problems concurrently. Use when facing 3+ independent failures or tasks.","ComeOnOliver\u002Fskillshub","ComeOnOliver","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F43215937?v=4","01M6E5JNFECJAAVFE0W7SMWWD2","Haoran Shu","01M6E5JNNDQS7AJZDYFQ8TPZCK",67,"MIT","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Faiskillstore\u002Fmarketplace\u002Fdmjgilbert\u002Fparallel-agents\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31],"parallel agents","parallel","tasks","work on","work","use","dispatch","dispatch multiple agents","2026-11-10 10:12:30",false,"---\nname: parallel-agents\ndescription: Dispatch multiple agents to work on independent problems concurrently. Use when facing 3+ independent failures or tasks.\n---\n\n# Dispatching Parallel Agents\n\nAssign separate agents to independent problem domains simultaneously for faster resolution.\n\n## When to Use\n\n- 3+ test failures across different files\u002Fsubsystems\n- Multiple independent tasks that don't share state\n- Investigations that won't interfere with each other\n- Failures from unrelated root causes\n\n## When NOT to Use\n\n- Failures are interconnected\n- Tasks share state or create conflicts\n- Agents would modify the same files\n- You lack context to properly scope tasks\n\n## Implementation Steps\n\n### 1. Group by Domain\n\nOrganize failures\u002Ftasks into independent categories:\n\n```markdown\nGroup A: Authentication tests (3 failures)\nGroup B: API endpoint tests (2 failures)\nGroup C: UI component tests (4 failures)\n```\n\n### 2. Define Focused Tasks\n\nEach agent receives:\n\n| Field       | Description                      |\n| ----------- | -------------------------------- |\n| Scope       | Specific files\u002Ftests to focus on |\n| Goal        | Clear success criteria           |\n| Constraints | What NOT to change               |\n| Output      | Expected deliverable             |\n\n### 3. Dispatch Concurrently\n\n**IMPORTANT**: Launch all tasks in a **single message** (no `run_in_background`). Multiple Task calls in the same message automatically run in parallel, and Claude waits for all to complete.\n\n```\n# All three tasks run in parallel automatically when in the same message\nTask(test-engineer, prompt=\"Fix auth test failures in src\u002Fauth\u002F*.test.ts\")\nTask(test-engineer, prompt=\"Fix API test failures in src\u002Fapi\u002F*.test.ts\")\nTask(frontend-developer, prompt=\"Fix UI test failures in src\u002Fcomponents\u002F*.test.tsx\")\n# Claude waits for all to complete, then continues\n```\n\n**Avoid `run_in_background: true`** unless you need to do other work while waiting. Task IDs must be captured and used within the same response.\n\n### 4. Integrate Results\n\n1. Review all agent outputs (available after parallel completion)\n2. Verify no conflicts between changes\n3. Run full test suite\n4. Merge changes\n\n## Effective Agent Prompts\n\n**Good prompt:**\n\n```\nFix the 3 failing tests in src\u002Fauth\u002Flogin.test.ts:\n- \"should reject invalid email format\"\n- \"should require password min length\"\n- \"should handle network errors\"\n\nError messages attached. Identify root causes - don't just increase timeouts.\nConstraints: Don't modify src\u002Fapi\u002F* files.\nOutput: Summary of fixes with test results.\n```\n\n**Bad prompt:**\n\n```\nFix all the tests\n```\n\n## Prompt Template\n\n```markdown\n## Task: [Specific description]\n\n**Scope:** [Files\u002Ftests to focus on]\n\n**Failures:**\n\n- [Test name]: [Error message]\n- [Test name]: [Error message]\n\n**Goal:** [What success looks like]\n\n**Constraints:**\n\n- Don't modify [files]\n- Preserve [behavior]\n\n**Output:**\n\n- Summary of root causes found\n- Changes made\n- Verification results\n```\n\n## Common Pitfalls\n\n| Mistake         | Problem                   | Solution               |\n| --------------- | ------------------------- | ---------------------- |\n| Vague scope     | Agent changes wrong files | Specify exact paths    |\n| Missing context | Agent can't diagnose      | Include error messages |\n| No constraints  | Conflicting changes       | Define boundaries      |\n| Unclear output  | Can't verify success      | Specify deliverables   |\n\n## Benefits\n\n- Reduces investigation time through parallelization\n- Each agent maintains narrow focus\n- Minimizes cross-agent interference\n- Solves multiple problems concurrently\n\n## Background Execution\n\nFor long-running tasks where you need to continue working, use `run_in_background: true`.\n\n### Pattern: Background + Foreground\n\n```\n# Long-running audit in background\naudit_task = Task(security-auditor,\n  prompt=\"Full security audit\",\n  run_in_background: true)\n\n# Continue with implementation work\nTask(frontend-developer, prompt=\"Build login form\")\n\n# Later, get audit results\nTaskOutput(audit_task.id, block: true)\n```\n\n### Pattern: Multiple Background Tasks\n\n```\n# Launch multiple background tasks\ntask1 = Task(test-engineer, prompt=\"...\", run_in_background: true)\ntask2 = Task(code-reviewer, prompt=\"...\", run_in_background: true)\n\n# Do other work...\n\n# Collect all results\nresult1 = TaskOutput(task1.id, block: true)\nresult2 = TaskOutput(task2.id, block: true)\n```\n\n### When to Use Background vs Foreground\n\n| Scenario                   | Mode                      | Why                          |\n| -------------------------- | ------------------------- | ---------------------------- |\n| Quick tasks (\u003C 1 min)      | Foreground                | Simpler, immediate results   |\n| Long audit\u002Fanalysis        | Background                | Continue working             |\n| Multiple independent tasks | Foreground (parallel)     | Auto-waits for all           |\n| Security + Implementation  | Background + Foreground   | Overlap work                 |\n\n### Important Notes\n\n- Task IDs are only valid within the same response\n- Always use `block: true` when retrieving results with TaskOutput\n- Prefer foreground parallel (single message, multiple Tasks) when possible\n- Background tasks should be collected before the response ends\n","# Dispatching Parallel Agents\n\nAssign separate agents to independent problem domains simultaneously for faster resolution.\n\n## When to Use\n\n- 3+ test failures across different files\u002Fsubsystems\n- Multiple independent tasks that don't share state\n- Investigations that won't interfere with each other\n- Failures from unrelated root causes\n\n## When NOT to Use\n\n- Failures are interconnected\n- Tasks share state or create conflicts\n- Agents would modify the same files\n- You lack context to properly scope tasks\n\n## Implementation Steps\n\n### 1. Group by Domain\n\nOrganize failures\u002Ftasks into independent categories:\n\n```markdown\nGroup A: Authentication tests (3 failures)\nGroup B: API endpoint tests (2 failures)\nGroup C: UI component tests (4 failures)\n```\n\n### 2. Define Focused Tasks\n\nEach agent receives:\n\n| Field       | Description                      |\n| ----------- | -------------------------------- |\n| Scope       | Specific files\u002Ftests to focus on |\n| Goal        | Clear success criteria           |\n| Constraints | What NOT to change               |\n| Output      | Expected deliverable             |\n\n### 3. Dispatch Concurrently\n\n**IMPORTANT**: Launch all tasks in a **single message** (no `run_in_background`). Multiple Task calls in the same message automatically run in parallel, and Claude waits for all to complete.\n\n```\n# All three tasks run in parallel automatically when in the same message\nTask(test-engineer, prompt=\"Fix auth test failures in src\u002Fauth\u002F*.test.ts\")\nTask(test-engineer, prompt=\"Fix API test failures in src\u002Fapi\u002F*.test.ts\")\nTask(frontend-developer, prompt=\"Fix UI test failures in src\u002Fcomponents\u002F*.test.tsx\")\n# Claude waits for all to complete, then continues\n```\n\n**Avoid `run_in_background: true`** unless you need to do other work while waiting. Task IDs must be captured and used within the same response.\n\n### 4. Integrate Results\n\n1. Review all agent outputs (available after parallel completion)\n2. Verify no conflicts between changes\n3. Run full test suite\n4. Merge changes\n\n## Effective Agent Prompts\n\n**Good prompt:**\n\n```\nFix the 3 failing tests in src\u002Fauth\u002Flogin.test.ts:\n- \"should reject invalid email format\"\n- \"should require password min length\"\n- \"should handle network errors\"\n\nError messages attached. Identify root causes - don't just increase timeouts.\nConstraints: Don't modify src\u002Fapi\u002F* files.\nOutput: Summary of fixes with test results.\n```\n\n**Bad prompt:**\n\n```\nFix all the tests\n```\n\n## Prompt Template\n\n```markdown\n## Task: [Specific description]\n\n**Scope:** [Files\u002Ftests to focus on]\n\n**Failures:**\n\n- [Test name]: [Error message]\n- [Test name]: [Error message]\n\n**Goal:** [What success looks like]\n\n**Constraints:**\n\n- Don't modify [files]\n- Preserve [behavior]\n\n**Output:**\n\n- Summary of root causes found\n- Changes made\n- Verification results\n```\n\n## Common Pitfalls\n\n| Mistake         | Problem                   | Solution               |\n| --------------- | ------------------------- | ---------------------- |\n| Vague scope     | Agent changes wrong files | Specify exact paths    |\n| Missing context | Agent can't diagnose      | Include error messages |\n| No constraints  | Conflicting changes       | Define boundaries      |\n| Unclear output  | Can't verify success      | Specify deliverables   |\n\n## Benefits\n\n- Reduces investigation time through parallelization\n- Each agent maintains narrow focus\n- Minimizes cross-agent interference\n- Solves multiple problems concurrently\n\n## Background Execution\n\nFor long-running tasks where you need to continue working, use `run_in_background: true`.\n\n### Pattern: Background + Foreground\n\n```\n# Long-running audit in background\naudit_task = Task(security-auditor,\n  prompt=\"Full security audit\",\n  run_in_background: true)\n\n# Continue with implementation work\nTask(frontend-developer, prompt=\"Build login form\")\n\n# Later, get audit results\nTaskOutput(audit_task.id, block: true)\n```\n\n### Pattern: Multiple Background Tasks\n\n```\n# Launch multiple background tasks\ntask1 = Task(test-engineer, prompt=\"...\", run_in_background: true)\ntask2 = Task(code-reviewer, prompt=\"...\", run_in_background: true)\n\n# Do other work...\n\n# Collect all results\nresult1 = TaskOutput(task1.id, block: true)\nresult2 = TaskOutput(task2.id, block: true)\n```\n\n### When to Use Background vs Foreground\n\n| Scenario                   | Mode                      | Why                          |\n| -------------------------- | ------------------------- | ---------------------------- |\n| Quick tasks (\u003C 1 min)      | Foreground                | Simpler, immediate results   |\n| Long audit\u002Fanalysis        | Background                | Continue working             |\n| Multiple independent tasks | Foreground (parallel)     | Auto-waits for all           |\n| Security + Implementation  | Background + Foreground   | Overlap work                 |\n\n### Important Notes\n\n- Task IDs are only valid within the same response\n- Always use `block: true` when retrieving results with TaskOutput\n- Prefer foreground parallel (single message, multiple Tasks) when possible\n- Background tasks should be collected before the response ends\n",{"name":5,"description":7},"main","def8531e65114c0fca8fb8551c1871ee0eed705c","skills\u002Faiskillstore\u002Fmarketplace\u002Fdmjgilbert\u002Fparallel-agents",24,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub","https:\u002F\u002Fskillshub.wtf","https:\u002F\u002Fgithub.com\u002FComeOnOliver","AI & Blockchain Engineer\r\n","2026-06-24 22:07:19",{"source":8,"skillName":5,"commands":47,"prompt":57,"zipAvailable":58},[48,51,54],{"tool":49,"command":50},"npx","npx skills add ComeOnOliver\u002Fskillshub --skill parallel-agents",{"tool":52,"command":53},"bunx","bunx skills add ComeOnOliver\u002Fskillshub --skill parallel-agents",{"tool":55,"command":56},"pnpm","pnpm dlx skills add ComeOnOliver\u002Fskillshub --skill parallel-agents","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Faiskillstore\u002Fmarketplace\u002Fdmjgilbert\u002Fparallel-agents\u002FSKILL.md",true,[60],{"path":61,"resourceType":62,"resourceTypeDesc":63,"sizeBytes":64,"sourceHtmlUrl":65},"skill-report.json","REFERENCE","参考文档",11238,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Faiskillstore\u002Fmarketplace\u002Fdmjgilbert\u002Fparallel-agents\u002Fskill-report.json",[67,77,87,96,102,109],{"skillNo":68,"name":69,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":73,"publishedAt":32},"01M6FKJN13GMG4KY57J3QD8GY5","model-hierarchy-skill","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzscole\u002Fmodel-hierarchy-skill\u002Fmodel-hierarchy-skill\u002FSKILL.md",[72],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[74,75,76],"model","model hierarchy","model hierarchy skill",{"skillNo":78,"name":79,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":83,"publishedAt":32},"01M6FKJJJ6GCAMR4PPPRD12M5X","zoom-mcp","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fzoom-mcp\u002FSKILL.md",[82],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[84,85,86],"MCP","zoom","zoom mcp",{"skillNo":88,"name":89,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":32},"01M6FKJE3KYT3KPEZMEEHYKA6K","zoom-apps-sdk","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fzoom-apps-sdk\u002FSKILL.md",[],[93,94,95,85],"sdk","zoom apps sdk","apps",{"skillNo":97,"name":98,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":99,"categories":100,"keywords":101,"publishedAt":32},"01M6FKJ9EDH9Q5VWDF8HAJRSP0","windows","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwindows\u002FSKILL.md",[],[98],{"skillNo":103,"name":104,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":105,"categories":106,"keywords":108,"publishedAt":32},"01M6FKJ666BEEFC3AE9GJ72Q93","whiteboard","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwhiteboard\u002FSKILL.md",[107],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[104],{"skillNo":110,"name":111,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":114,"publishedAt":32},"01M6FKJ21WFCDGTJQ2Q91T9DVW","websockets","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwebsockets\u002FSKILL.md",[],[111]]