[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DFVB3RD4MCQTHZ3T55PT6D":3,"skill-same:01M6DFVB3RD4MCQTHZ3T55PT6D":90},{"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":18,"publishedAt":31,"linkOnly":32,"markdownContent":33,"markdownBody":34,"frontMatter":35,"refName":44,"commitSha":45,"skillDirectory":46,"forksCount":47,"repositoryHtmlUrl":48,"repositoryHomepageUrl":6,"ownerHtmlUrl":49,"ownerBlogUrl":6,"ownerDescription":50,"updatedAt":51,"install":52,"resources":65},"01M6DFVB3RD4MCQTHZ3T55PT6D","delegation-tuning",null,"Tune Hermes delegation for parallel subagent execution.","zapabob\u002FSkills","zapabob","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F130830376?v=4","01M6DBMYZAGRTFQDSQYW3W764P","峯岸 亮","01M6DBMZ3GKDSJ01NZ9XTZD7JD",2,"MIT","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002F53\u002F53761ea4905fd5a4087df6a6d5ec49a2d1f3edb64d138a09a18e5b4c4c63eaad\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30],"delegation","tuning","delegation tuning","macos","linux","windows","execution","subagent","parallel","hermes","hermes agent","devops","2026-11-10 10:12:30",false,"---\r\nname: delegation-tuning\r\ndescription: Tune Hermes delegation for parallel subagent execution.\r\nversion: 1.0.0\r\nauthor: Hermes Agent\r\nlicense: MIT\r\nplatforms: [linux, macos, windows]\r\nmetadata.hermes.category: devops\r\nmetadata.hermes.config:\r\n  delegation.max_concurrent_children: int\r\n  delegation.provider: string\r\n  delegation.model: string\r\n  delegation.max_spawn_depth: int\r\n  delegation.orchestrator_enabled: bool\r\n  delegation.subagent_auto_approve: bool\r\n  delegation.child_timeout_seconds: int or null\r\n---\r\n\r\n## When to Use\r\nUse this skill when you need to run multiple subagents in parallel via `delegate_task` and want to optimize throughput, reduce latency, or avoid hitting provider rate limits. Typical scenarios include:\r\n- Running independent research tasks concurrently.\r\n- Parallel code review or linting across multiple files.\r\n- Batch processing of data where each chunk can be handled by a subagent.\r\n\r\n## Prerequisites\r\n- Hermes Agent v0.19.0 or later.\r\n- A working LLM provider configured for subagents (see `delegation.provider` and `delegation.model`).\r\n- Sufficient quota\u002Frate limits on the chosen provider to handle increased concurrent requests.\r\n\r\n## How to Run\r\nAdjust the delegation settings in `~\u002F.hermes\u002Fconfig.yaml` (or via `hermes config set`) then run your delegated task as usual. No special command is needed; the settings apply automatically to any `delegate_task` call.\r\n\r\n## Procedure\r\n1. **Set concurrency level**\r\n   Increase `delegation.max_concurrent_children` to the desired number of parallel subagents (default 3).\r\n   Example: `hermes config set delegation.max_concurrent_children 4`\r\n\r\n2. **Choose provider and model for subagents**\r\n   By default subagents inherit the parent's provider\u002Fmodel. To override, set:\r\n   - `delegation.provider` (e.g., `openai`, `nous`, `anthropic`)\r\n   - `delegation.model` (e.g., `qwen3`, `gpt-4o-mini`)\r\n   Leave blank to inherit.\r\n\r\n3. **Control nesting depth**\r\n   Set `delegation.max_spawn_depth` to allow orchestrator subagents to spawn their own workers.\r\n   - `1` (default): flat tree – only the parent can spawn workers.\r\n   - `2` or higher: enables nested orchestration.\r\n\r\n4. **Enable orchestrator role**\r\n   Ensure `delegation.orchestrator_enabled` is `true` (default) if you plan to use `role: orchestrator` in your delegation calls.\r\n\r\n5. **Optionally allow auto‑approval of dangerous commands in subagents**\r\n   Set `delegation.subagent_auto_approve: true` only if you trust the subagent's actions and want to skip approval prompts (use with caution).\r\n\r\n6. **Set a wall‑clock timeout per subagent (optional)**\r\n   Configure `delegation.child_timeout_seconds` (seconds) to kill stalled subagents.\r\n   Set to `null` or `0` to disable (default).\r\n\r\n7. **Test the configuration**\r\n   Run a simple parallel delegation to verify behavior, e.g.:\r\n   ```\r\n   hermes -z \"Use delegate_task to run four tasks in parallel: each should echo its number (1‑4) and sleep 1 second. Return the results as a list.\"\r\n   ```\r\n   Observe that total time approximates the longest child runtime (≈1 s + overhead) rather than the sum.\r\n\r\n## Verification\r\n- Check that the number of concurrent subagent processes matches `delegation.max_concurrent_children` (visible via `hermes process list` or the delegation spinner in the TUI).\r\n- Confirm that child agent logs show the expected provider\u002Fmodel (if overridden).\r\n- Ensure that increasing `max_concurrent_children` reduces overall wall‑time for embarrassingly parallel workloads, up to the point of provider rate‑limit or resource saturation.\r\n- If you see frequent `child_timeout` errors, increase `child_timeout_seconds` or investigate the subagent's workload for stalls.\r\n\r\n## Pitfalls\r\n- Raising `max_concurrent_children` too high can cause provider‑side rate‑limit errors (HTTP 429). Monitor logs and back off if needed.\r\n- Setting `delegation.model` to a model not supported by the chosen provider will cause fallback errors; verify compatibility.\r\n- Enabling `subagent_auto_approve` can let subagents run destructive commands without oversight; only enable in trusted, sandboxed contexts.\r\n- Nested delegation (`max_spawn_depth` > 1) increases API usage multiplicatively; plan quotas accordingly.\r\n- **Delegation capture failure**: If you see repeated `Delegation owner exited before recording a terminal result; outcome unknown` errors across multiple jobs, the delegation layer may be broken. In this case, bypass delegation entirely by running scripts directly in foreground mode using `terminal(command=\"py -3 script.py\", workdir=\"path\u002Fto\u002Fscript\")` to capture full output and exit codes.\r\n\r\n## Debugging Delegation Failures\r\nWhen delegation consistently fails to capture results (showing `status=unknown` or `Delegation owner exited before recording a terminal result`):\r\n\r\n1. **Pause affected cron jobs** to stop the delegation loop:\r\n   ```bash\r\n   cronjob action=pause job_id=\u003Cjob_id>\r\n   ```\r\n\r\n2. **Run the script directly in foreground mode** to capture actual output:\r\n   ```bash\r\n   terminal command=\"py -3 C:\\\\path\\\\to\\\\script.py\" timeout=180 workdir=\"C:\\\\path\\\\to\\\\script\\\\dir\"\r\n   ```\r\n\r\n3. **Examine the script directly** to identify failure points:\r\n   ```bash\r\n   read_file path=\"C:\\\\path\\\\to\\\\script.py\"\r\n   ```\r\n\r\n4. **Fix identified issues** (OAuth tokens, locks, timeouts, etc.) then resume the cron job.\r\n\r\nThis approach bypasses the broken delegation capture mechanism and provides reliable debugging visibility.\r\n\r\n## Quick Reference\r\n| Setting | Example | Effect |\r\n|---------|---------|--------|\r\n| `delegation.max_concurrent_children` | `4` | Number of subagents that can run in parallel per delegation call. |\r\n| `delegation.provider` | `openai` | LLM provider used by subagents (blank = inherit). |\r\n| `delegation.model` | `qwen3` | Model name used by subagents (blank = inherit). |\r\n| `delegation.max_spawn_depth` | `2` | Allows orchestrator subagents to spawn their own workers. |\r\n| `delegation.orchestrator_enabled` | `true` | Enables the `orchestrator` role for subagents. |\r\n| `delegation.subagent_auto_approve` | `false` | When `true`, skips dangerous‑command prompts in subagents. |\r\n| `delegation.child_timeout_seconds` | `120` | Hard wall‑clock limit per subagent (seconds); `null` disables. |","## When to Use\r\nUse this skill when you need to run multiple subagents in parallel via `delegate_task` and want to optimize throughput, reduce latency, or avoid hitting provider rate limits. Typical scenarios include:\r\n- Running independent research tasks concurrently.\r\n- Parallel code review or linting across multiple files.\r\n- Batch processing of data where each chunk can be handled by a subagent.\r\n\r\n## Prerequisites\r\n- Hermes Agent v0.19.0 or later.\r\n- A working LLM provider configured for subagents (see `delegation.provider` and `delegation.model`).\r\n- Sufficient quota\u002Frate limits on the chosen provider to handle increased concurrent requests.\r\n\r\n## How to Run\r\nAdjust the delegation settings in `~\u002F.hermes\u002Fconfig.yaml` (or via `hermes config set`) then run your delegated task as usual. No special command is needed; the settings apply automatically to any `delegate_task` call.\r\n\r\n## Procedure\r\n1. **Set concurrency level**\r\n   Increase `delegation.max_concurrent_children` to the desired number of parallel subagents (default 3).\r\n   Example: `hermes config set delegation.max_concurrent_children 4`\r\n\r\n2. **Choose provider and model for subagents**\r\n   By default subagents inherit the parent's provider\u002Fmodel. To override, set:\r\n   - `delegation.provider` (e.g., `openai`, `nous`, `anthropic`)\r\n   - `delegation.model` (e.g., `qwen3`, `gpt-4o-mini`)\r\n   Leave blank to inherit.\r\n\r\n3. **Control nesting depth**\r\n   Set `delegation.max_spawn_depth` to allow orchestrator subagents to spawn their own workers.\r\n   - `1` (default): flat tree – only the parent can spawn workers.\r\n   - `2` or higher: enables nested orchestration.\r\n\r\n4. **Enable orchestrator role**\r\n   Ensure `delegation.orchestrator_enabled` is `true` (default) if you plan to use `role: orchestrator` in your delegation calls.\r\n\r\n5. **Optionally allow auto‑approval of dangerous commands in subagents**\r\n   Set `delegation.subagent_auto_approve: true` only if you trust the subagent's actions and want to skip approval prompts (use with caution).\r\n\r\n6. **Set a wall‑clock timeout per subagent (optional)**\r\n   Configure `delegation.child_timeout_seconds` (seconds) to kill stalled subagents.\r\n   Set to `null` or `0` to disable (default).\r\n\r\n7. **Test the configuration**\r\n   Run a simple parallel delegation to verify behavior, e.g.:\r\n   ```\r\n   hermes -z \"Use delegate_task to run four tasks in parallel: each should echo its number (1‑4) and sleep 1 second. Return the results as a list.\"\r\n   ```\r\n   Observe that total time approximates the longest child runtime (≈1 s + overhead) rather than the sum.\r\n\r\n## Verification\r\n- Check that the number of concurrent subagent processes matches `delegation.max_concurrent_children` (visible via `hermes process list` or the delegation spinner in the TUI).\r\n- Confirm that child agent logs show the expected provider\u002Fmodel (if overridden).\r\n- Ensure that increasing `max_concurrent_children` reduces overall wall‑time for embarrassingly parallel workloads, up to the point of provider rate‑limit or resource saturation.\r\n- If you see frequent `child_timeout` errors, increase `child_timeout_seconds` or investigate the subagent's workload for stalls.\r\n\r\n## Pitfalls\r\n- Raising `max_concurrent_children` too high can cause provider‑side rate‑limit errors (HTTP 429). Monitor logs and back off if needed.\r\n- Setting `delegation.model` to a model not supported by the chosen provider will cause fallback errors; verify compatibility.\r\n- Enabling `subagent_auto_approve` can let subagents run destructive commands without oversight; only enable in trusted, sandboxed contexts.\r\n- Nested delegation (`max_spawn_depth` > 1) increases API usage multiplicatively; plan quotas accordingly.\r\n- **Delegation capture failure**: If you see repeated `Delegation owner exited before recording a terminal result; outcome unknown` errors across multiple jobs, the delegation layer may be broken. In this case, bypass delegation entirely by running scripts directly in foreground mode using `terminal(command=\"py -3 script.py\", workdir=\"path\u002Fto\u002Fscript\")` to capture full output and exit codes.\r\n\r\n## Debugging Delegation Failures\r\nWhen delegation consistently fails to capture results (showing `status=unknown` or `Delegation owner exited before recording a terminal result`):\r\n\r\n1. **Pause affected cron jobs** to stop the delegation loop:\r\n   ```bash\r\n   cronjob action=pause job_id=\u003Cjob_id>\r\n   ```\r\n\r\n2. **Run the script directly in foreground mode** to capture actual output:\r\n   ```bash\r\n   terminal command=\"py -3 C:\\\\path\\\\to\\\\script.py\" timeout=180 workdir=\"C:\\\\path\\\\to\\\\script\\\\dir\"\r\n   ```\r\n\r\n3. **Examine the script directly** to identify failure points:\r\n   ```bash\r\n   read_file path=\"C:\\\\path\\\\to\\\\script.py\"\r\n   ```\r\n\r\n4. **Fix identified issues** (OAuth tokens, locks, timeouts, etc.) then resume the cron job.\r\n\r\nThis approach bypasses the broken delegation capture mechanism and provides reliable debugging visibility.\r\n\r\n## Quick Reference\r\n| Setting | Example | Effect |\r\n|---------|---------|--------|\r\n| `delegation.max_concurrent_children` | `4` | Number of subagents that can run in parallel per delegation call. |\r\n| `delegation.provider` | `openai` | LLM provider used by subagents (blank = inherit). |\r\n| `delegation.model` | `qwen3` | Model name used by subagents (blank = inherit). |\r\n| `delegation.max_spawn_depth` | `2` | Allows orchestrator subagents to spawn their own workers. |\r\n| `delegation.orchestrator_enabled` | `true` | Enables the `orchestrator` role for subagents. |\r\n| `delegation.subagent_auto_approve` | `false` | When `true`, skips dangerous‑command prompts in subagents. |\r\n| `delegation.child_timeout_seconds` | `120` | Hard wall‑clock limit per subagent (seconds); `null` disables. |",{"name":5,"author":36,"license":15,"version":37,"platforms":38,"description":7,"metadata.hermes.config":39,"metadata.hermes.category":30},"Hermes Agent","1.0.0",[23,22,24],{"delegation.model":40,"delegation.provider":40,"delegation.max_spawn_depth":41,"delegation.orchestrator_enabled":42,"delegation.child_timeout_seconds":43,"delegation.subagent_auto_approve":42,"delegation.max_concurrent_children":41},"string","int","bool","int or null","main","0d1434f89f4bc6b07e9d272accd8c1013768820f","archive\u002Fskill-md-reviewed\u002Fsha256\u002F53\u002F53761ea4905fd5a4087df6a6d5ec49a2d1f3edb64d138a09a18e5b4c4c63eaad",0,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills","https:\u002F\u002Fgithub.com\u002Fzapabob","The Open University of JAPAN　2019-now\r\n\r\nTokyo University of Science　2007-2010","2026-10-26 16:29:01",{"source":8,"skillName":5,"commands":53,"prompt":63,"zipAvailable":64},[54,57,60],{"tool":55,"command":56},"npx","npx skills add zapabob\u002FSkills --skill delegation-tuning",{"tool":58,"command":59},"bunx","bunx skills add zapabob\u002FSkills --skill delegation-tuning",{"tool":61,"command":62},"pnpm","pnpm dlx skills add zapabob\u002FSkills --skill delegation-tuning","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002F53\u002F53761ea4905fd5a4087df6a6d5ec49a2d1f3edb64d138a09a18e5b4c4c63eaad\u002FSKILL.md",true,[66,72,76,82,86],{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"assets\u002Fmodule-help.csv","ASSET","资产文件",112,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.gemini\u002Fskills\u002Fbmad-module-builder\u002Fassets\u002Fsetup-skill-template\u002Fassets\u002Fmodule-help.csv",{"path":73,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":74,"sourceHtmlUrl":75},"assets\u002Fmodule.yaml",96,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.gemini\u002Fskills\u002Fbmad-module-builder\u002Fassets\u002Fsetup-skill-template\u002Fassets\u002Fmodule.yaml",{"path":77,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":80,"sourceHtmlUrl":81},"scripts\u002Fcleanup-legacy.py","SCRIPT","脚本",7743,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.gemini\u002Fskills\u002Fbmad-module-builder\u002Fassets\u002Fsetup-skill-template\u002Fscripts\u002Fcleanup-legacy.py",{"path":83,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":84,"sourceHtmlUrl":85},"scripts\u002Fmerge-config.py",14139,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.gemini\u002Fskills\u002Fbmad-module-builder\u002Fassets\u002Fsetup-skill-template\u002Fscripts\u002Fmerge-config.py",{"path":87,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":88,"sourceHtmlUrl":89},"scripts\u002Fmerge-help-csv.py",6607,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.gemini\u002Fskills\u002Fbmad-module-builder\u002Fassets\u002Fsetup-skill-template\u002Fscripts\u002Fmerge-help-csv.py",[91,124,158,185,211,227],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":110,"publishedAt":31},"01M6GJTE1GESGGF9K028Y7753E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fxurl\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[97,102,105],{"categoryNo":98,"categoryName":99,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000010","后端与接口","SCENARIO","使用场景",{"categoryNo":103,"categoryName":104,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":106,"categoryName":107,"taxonomyType":108,"taxonomyTypeDesc":109},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[93,111,22,112,113,23,24,114,115,116,117,118,119,28,29,120,44,121,122,123],"openclaw","cli","github","official","search","https","social-media","api","Media","twitter","social","via","post",{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":128,"categories":129,"keywords":141,"publishedAt":31},"01M6GJT8S1PXQ8WPXT1HSNBWH2","writing-plans","Write implementation plans: bite-sized tasks, paths, code.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwriting-plans\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[130,133,136,137,138],{"categoryNo":131,"categoryName":132,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":134,"categoryName":135,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":103,"categoryName":104,"taxonomyType":100,"taxonomyTypeDesc":101},{"categoryNo":106,"categoryName":107,"taxonomyType":108,"taxonomyTypeDesc":109},{"categoryNo":139,"categoryName":140,"taxonomyType":108,"taxonomyTypeDesc":109},"01JBSEEDCAT000000000000036","内容创作者",[142,143,144,145,22,146,23,147,24,148,149,150,151,152,153,26,154,155,156,157],"writing","writing plans","plans","design","review","workflow","documentation","tasks","development","code","planning","code-review","write","superpowers","paths","test-driven-development",{"skillNo":159,"name":160,"nameZh":6,"description":161,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":162,"categories":163,"keywords":167,"publishedAt":31},"01M6GJST0H5X7WHBJ3BZK0DVKH","whisper","OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwhisper\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[164],{"categoryNo":165,"categoryName":166,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000016","图像与多媒体",[160,22,23,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184],"openai","speech","multimodal","Transformers","ASR","Audio","text","model","speech recognition","use","general","multilingual","translation","english","identification","processing","Audio Processing",{"skillNo":186,"name":187,"nameZh":6,"description":188,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":189,"categories":190,"keywords":194,"publishedAt":31},"01M6GJSMRQ8EETZTYMG72P70EA","weights-and-biases","W&B: log ML experiments, sweeps, model registry, dashboards.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fweights-and-biases\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[191],{"categoryNo":192,"categoryName":193,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000013","分析与可视化",[195,22,196,197,24,198,199,200,201,20,202,175,203,204,205,206,207,208,209,210],"weights and biases","tensorflow","experiments","pytorch","visualization","registry","real-time","dashboards","experiment","log","time","Experiment Tracking","Model Registry","Hyperparameter Tuning","WandB","Real-Time Visualization",{"skillNo":212,"name":213,"nameZh":6,"description":214,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":215,"categories":216,"keywords":218,"publishedAt":31},"01M6GJS19KDPRY112JTR326S06","watchers","Poll RSS, JSON APIs, and GitHub with watermark dedup.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwatchers\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[217],{"categoryNo":103,"categoryName":104,"taxonomyType":100,"taxonomyTypeDesc":101},[213,22,219,113,23,220,221,222,223,224,225,28,29,30,226],"terminal","automation","http","monitoring","cron","json","rss","with",{"skillNo":228,"name":229,"nameZh":6,"description":230,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":231,"categories":232,"keywords":235,"publishedAt":31},"01M6GJRHJD2A6Q830SWF9W48BC","serving-llms-vllm","vLLM: high-throughput LLM serving, OpenAI API, quantization.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fvllm\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[233,234],{"categoryNo":98,"categoryName":99,"taxonomyType":100,"taxonomyTypeDesc":101},{"categoryNo":106,"categoryName":107,"taxonomyType":108,"taxonomyTypeDesc":109},[236,237,238,23,239,168,240,241,242,171,243,118,244,245,246,247,248,249,250,251],"llms","vLLM","serving llms vllm","research","openai api","tensor","Inference Serving","Production","llm serving","High Throughput","PagedAttention","Continuous Batching","Quantization","Tensor Parallelism","high-throughput","throughput"]