[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DG6JAXB8PQSW0XNR30ZCSR":3,"skill-same:01M6DG6JAXB8PQSW0XNR30ZCSR":80},{"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":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":52,"commitSha":53,"skillDirectory":54,"forksCount":55,"repositoryHtmlUrl":56,"repositoryHomepageUrl":6,"ownerHtmlUrl":57,"ownerBlogUrl":6,"ownerDescription":58,"updatedAt":59,"install":60,"resources":73},"01M6DG6JAXB8PQSW0XNR30ZCSR","delegation-fallback",null,"Delegated worker failed: fall back to direct 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\u002F5a\u002F5ad9451b48bb95325402d6f6b006173bcd5360bf24f83822835a14d5a68c9f4a\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35],"delegation fallback","fallback","delegation","execution","OpenManus","worker","back","runner","hermes","direct","macos","windows","web","automation","research","codex","linux","2026-11-10 10:12:30",false,"---\r\nname: delegation-fallback\r\ndescription: \"Delegated worker failed: fall back to direct execution.\"\r\nversion: 1.0.0\r\nauthor: Hermes\r\nlicense: MIT\r\nplatforms: [linux, macos, windows]\r\nmetadata:\r\n  hermes:\r\n    tags: [delegation, fallback, research, execution, openmanus, codex]\r\n    related_skills: [openmanus-delegation, web-fallback, research-automation-runner]\r\n---\r\n\r\n# Delegation Fallback\r\n\r\nWhen a delegated worker fails at runtime (not at planning), **do not block on the user**. Fall back to direct execution using the agent's own tools. This pattern applies to OpenManus, Codex, Claude Code, and any subagent\u002Fruntime that can fail independently.\r\n\r\n## Trigger\r\n\r\n- Worker exits non-zero with `Connection error`, `AuthenticationError`, `APIConnectionError`\r\n- Worker times out at the 420s tool cap with no receipt\r\n- Worker fails before first LLM call (missing runtime, bad config, unreachable endpoint)\r\n- Worker reports `status: failed` with no `receipt.json`\r\n\r\n## Core Rule\r\n\r\n**A worker failure is not a task failure.** The agent that spawned the worker has its own tools (`web_search`, `web_extract`, `terminal`, `execute_code`, `read_file`). When the worker can't run, do the work directly and report what you produced. Only stop for true blockers (irreversible actions, secrets, explicit user gates).\r\n\r\n## Fallback Patterns\r\n\r\n### OpenManus worker fails (API key \u002F runtime \u002F network)\r\n\r\n1. Diagnose from stderr:\r\n   - `AuthenticationError` \u002F 401 → `api_key_env` points to invalid key in `.hermes\u002F.env`\r\n   - `Connection error` → LLM endpoint unreachable (llama-server down, wrong URL)\r\n   - `failed to remove directory \\.venv\\Lib` → `runtime_command` not set, `uv run` tried to rebuild repo venv\r\n   - `error: every wide-research item needs a non-empty prompt` → wrong `items` schema (object vs string)\r\n2. If fixable in \u003C30s (set `runtime_command`, restart llama-server), fix and retry ONCE.\r\n3. If not fixable (no API key, llama-server binary missing), **fall back to direct research**:\r\n   - `web_search` with multiple query angles (4-8 parallel queries)\r\n   - `web_extract` on the top URLs (char_limit=12000-15000 per page)\r\n   - `execute_code` to parse\u002Fmerge results\r\n   - Synthesize a report from the extracted content\r\n4. Report: \"OpenManus は [reason] で稼働できなかったため、直接リサーチしました。\" + the report.\r\n\r\n### Codex \u002F Claude Code CLI fails (auth \u002F timeout \u002F sandbox)\r\n\r\n1. Check `stderr` for auth vs timeout vs sandbox denial.\r\n2. If auth: suggest `codex login` \u002F `claude login` to user, fall back to direct execution meanwhile.\r\n3. If timeout: the task is too big for the CLI's context; decompose and run directly.\r\n4. If sandbox denial: preserve the sandbox and authorization boundary. Use an\r\n   already authorized equivalent within scope, or request approval if scope must\r\n   change. Never bypass a denied operation by switching tools.\r\n\r\n### Subagent (delegate_task) fails\r\n\r\n1. Check the child's final summary for the failure reason.\r\n2. If the child hit a tool error (not a reasoning failure), retry the task directly in the parent using the same tools the child would have used.\r\n3. If the child's reasoning was flawed, steer it via `delegate_task(action='steer')` rather than abandoning it.\r\n\r\n## When NOT to Fall Back\r\n\r\n- The task requires the worker's specific capabilities (e.g., Daytona sandbox, browser automation with persistent login state) AND no equivalent exists in the parent.\r\n- The user explicitly asked to wait for the worker.\r\n- The failure is in the worker's reasoning (not infrastructure) — steer or restart instead.\r\n\r\n## Verification\r\n\r\nAfter fallback delivery:\r\n- Confirm the report covers the same topics the worker was asked about.\r\n- Cite sources (URLs) so the user can verify.\r\n- Note any gaps the worker might have filled (e.g., \"OpenManus would have browsed interactively; this is static extraction\").\r\n\r\n## Pitfalls\r\n\r\n- **Don't retry the same failing worker 3x.** One retry after a fix attempt; then fall back.\r\n- **Don't pretend the worker ran.** Be transparent: \"Worker failed, did it directly.\"\r\n- **Don't block on user input for fixable issues.** Fix and retry once; only ask if the fix needs secrets or irreversible changes.\r\n- **420s cap is not a failure.** If the worker is still running past 420s, read its `receipt.json` — it may have completed normally.\r\n","# Delegation Fallback\r\n\r\nWhen a delegated worker fails at runtime (not at planning), **do not block on the user**. Fall back to direct execution using the agent's own tools. This pattern applies to OpenManus, Codex, Claude Code, and any subagent\u002Fruntime that can fail independently.\r\n\r\n## Trigger\r\n\r\n- Worker exits non-zero with `Connection error`, `AuthenticationError`, `APIConnectionError`\r\n- Worker times out at the 420s tool cap with no receipt\r\n- Worker fails before first LLM call (missing runtime, bad config, unreachable endpoint)\r\n- Worker reports `status: failed` with no `receipt.json`\r\n\r\n## Core Rule\r\n\r\n**A worker failure is not a task failure.** The agent that spawned the worker has its own tools (`web_search`, `web_extract`, `terminal`, `execute_code`, `read_file`). When the worker can't run, do the work directly and report what you produced. Only stop for true blockers (irreversible actions, secrets, explicit user gates).\r\n\r\n## Fallback Patterns\r\n\r\n### OpenManus worker fails (API key \u002F runtime \u002F network)\r\n\r\n1. Diagnose from stderr:\r\n   - `AuthenticationError` \u002F 401 → `api_key_env` points to invalid key in `.hermes\u002F.env`\r\n   - `Connection error` → LLM endpoint unreachable (llama-server down, wrong URL)\r\n   - `failed to remove directory \\.venv\\Lib` → `runtime_command` not set, `uv run` tried to rebuild repo venv\r\n   - `error: every wide-research item needs a non-empty prompt` → wrong `items` schema (object vs string)\r\n2. If fixable in \u003C30s (set `runtime_command`, restart llama-server), fix and retry ONCE.\r\n3. If not fixable (no API key, llama-server binary missing), **fall back to direct research**:\r\n   - `web_search` with multiple query angles (4-8 parallel queries)\r\n   - `web_extract` on the top URLs (char_limit=12000-15000 per page)\r\n   - `execute_code` to parse\u002Fmerge results\r\n   - Synthesize a report from the extracted content\r\n4. Report: \"OpenManus は [reason] で稼働できなかったため、直接リサーチしました。\" + the report.\r\n\r\n### Codex \u002F Claude Code CLI fails (auth \u002F timeout \u002F sandbox)\r\n\r\n1. Check `stderr` for auth vs timeout vs sandbox denial.\r\n2. If auth: suggest `codex login` \u002F `claude login` to user, fall back to direct execution meanwhile.\r\n3. If timeout: the task is too big for the CLI's context; decompose and run directly.\r\n4. If sandbox denial: preserve the sandbox and authorization boundary. Use an\r\n   already authorized equivalent within scope, or request approval if scope must\r\n   change. Never bypass a denied operation by switching tools.\r\n\r\n### Subagent (delegate_task) fails\r\n\r\n1. Check the child's final summary for the failure reason.\r\n2. If the child hit a tool error (not a reasoning failure), retry the task directly in the parent using the same tools the child would have used.\r\n3. If the child's reasoning was flawed, steer it via `delegate_task(action='steer')` rather than abandoning it.\r\n\r\n## When NOT to Fall Back\r\n\r\n- The task requires the worker's specific capabilities (e.g., Daytona sandbox, browser automation with persistent login state) AND no equivalent exists in the parent.\r\n- The user explicitly asked to wait for the worker.\r\n- The failure is in the worker's reasoning (not infrastructure) — steer or restart instead.\r\n\r\n## Verification\r\n\r\nAfter fallback delivery:\r\n- Confirm the report covers the same topics the worker was asked about.\r\n- Cite sources (URLs) so the user can verify.\r\n- Note any gaps the worker might have filled (e.g., \"OpenManus would have browsed interactively; this is static extraction\").\r\n\r\n## Pitfalls\r\n\r\n- **Don't retry the same failing worker 3x.** One retry after a fix attempt; then fall back.\r\n- **Don't pretend the worker ran.** Be transparent: \"Worker failed, did it directly.\"\r\n- **Don't block on user input for fixable issues.** Fix and retry once; only ask if the fix needs secrets or irreversible changes.\r\n- **420s cap is not a failure.** If the worker is still running past 420s, read its `receipt.json` — it may have completed normally.\r\n",{"name":5,"author":41,"license":15,"version":42,"metadata":43,"platforms":51,"description":7},"Hermes","1.0.0",{"hermes":44},{"tags":45,"related_skills":47},[21,20,33,22,46,34],"openmanus",[48,49,50],"openmanus-delegation","web-fallback","research-automation-runner",[35,29,30],"main","0d1434f89f4bc6b07e9d272accd8c1013768820f","archive\u002Fskill-md-reviewed\u002Fsha256\u002F5a\u002F5ad9451b48bb95325402d6f6b006173bcd5360bf24f83822835a14d5a68c9f4a",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":61,"prompt":71,"zipAvailable":72},[62,65,68],{"tool":63,"command":64},"npx","npx skills add zapabob\u002FSkills --skill delegation-fallback",{"tool":66,"command":67},"bunx","bunx skills add zapabob\u002FSkills --skill delegation-fallback",{"tool":69,"command":70},"pnpm","pnpm dlx skills add zapabob\u002FSkills --skill delegation-fallback","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002F5a\u002F5ad9451b48bb95325402d6f6b006173bcd5360bf24f83822835a14d5a68c9f4a\u002FSKILL.md",true,[74],{"path":75,"resourceType":76,"resourceTypeDesc":77,"sizeBytes":78,"sourceHtmlUrl":79},"index.json","REFERENCE","参考文档",531,"https:\u002F\u002Fgithub.com\u002FDCx7C5\u002Fai-marketplace\u002Fblob\u002Fmain\u002Fskills\u002Fcloud-aws-guardduty\u002Findex.json",[81,115,150,177,204,220],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":100,"publishedAt":36},"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",[87,92,95],{"categoryNo":88,"categoryName":89,"taxonomyType":90,"taxonomyTypeDesc":91},"01JBSEEDCAT000000000000010","后端与接口","SCENARIO","使用场景",{"categoryNo":93,"categoryName":94,"taxonomyType":90,"taxonomyTypeDesc":91},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":96,"categoryName":97,"taxonomyType":98,"taxonomyTypeDesc":99},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[83,101,29,102,103,35,30,104,105,106,107,108,109,27,110,111,52,112,113,114],"openclaw","cli","github","official","search","https","social-media","api","Media","hermes agent","twitter","social","via","post",{"skillNo":116,"name":117,"nameZh":6,"description":118,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":119,"categories":120,"keywords":132,"publishedAt":36},"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",[121,124,127,128,129],{"categoryNo":122,"categoryName":123,"taxonomyType":90,"taxonomyTypeDesc":91},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":125,"categoryName":126,"taxonomyType":90,"taxonomyTypeDesc":91},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":93,"categoryName":94,"taxonomyType":90,"taxonomyTypeDesc":91},{"categoryNo":96,"categoryName":97,"taxonomyType":98,"taxonomyTypeDesc":99},{"categoryNo":130,"categoryName":131,"taxonomyType":98,"taxonomyTypeDesc":99},"01JBSEEDCAT000000000000036","内容创作者",[133,134,135,136,29,137,35,138,30,139,140,141,142,143,144,145,146,147,148,149],"writing","writing plans","plans","design","review","workflow","documentation","tasks","development","code","planning","code-review","subagent","write","superpowers","paths","test-driven-development",{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":159,"publishedAt":36},"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",[156],{"categoryNo":157,"categoryName":158,"taxonomyType":90,"taxonomyTypeDesc":91},"01JBSEEDCAT000000000000016","图像与多媒体",[152,29,35,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176],"openai","speech","multimodal","Transformers","ASR","Audio","text","model","speech recognition","use","general","multilingual","translation","english","identification","processing","Audio Processing",{"skillNo":178,"name":179,"nameZh":6,"description":180,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":181,"categories":182,"keywords":186,"publishedAt":36},"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",[183],{"categoryNo":184,"categoryName":185,"taxonomyType":90,"taxonomyTypeDesc":91},"01JBSEEDCAT000000000000013","分析与可视化",[187,29,188,189,30,190,191,192,193,194,195,167,196,197,198,199,200,201,202,203],"weights and biases","tensorflow","experiments","pytorch","visualization","registry","real-time","tuning","dashboards","experiment","log","time","Experiment Tracking","Model Registry","Hyperparameter Tuning","WandB","Real-Time Visualization",{"skillNo":205,"name":206,"nameZh":6,"description":207,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":208,"categories":209,"keywords":211,"publishedAt":36},"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",[210],{"categoryNo":93,"categoryName":94,"taxonomyType":90,"taxonomyTypeDesc":91},[206,29,212,103,35,32,213,214,215,216,217,27,110,218,219],"terminal","http","monitoring","cron","json","rss","devops","with",{"skillNo":221,"name":222,"nameZh":6,"description":223,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":224,"categories":225,"keywords":228,"publishedAt":36},"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",[226,227],{"categoryNo":88,"categoryName":89,"taxonomyType":90,"taxonomyTypeDesc":91},{"categoryNo":96,"categoryName":97,"taxonomyType":98,"taxonomyTypeDesc":99},[229,230,231,35,33,160,232,233,234,163,235,108,236,237,238,239,240,241,242,243],"llms","vLLM","serving llms vllm","openai api","tensor","Inference Serving","Production","llm serving","High Throughput","PagedAttention","Continuous Batching","Quantization","Tensor Parallelism","high-throughput","throughput"]