[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6G9K2WJDDFPRYJ2668A9NMP":3,"skill-same:01M6G9K2WJDDFPRYJ2668A9NMP":71},{"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":34,"linkOnly":35,"markdownContent":36,"markdownBody":37,"frontMatter":38,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":52,"repositoryHtmlUrl":53,"repositoryHomepageUrl":6,"ownerHtmlUrl":54,"ownerBlogUrl":6,"ownerDescription":55,"updatedAt":56,"install":57,"resources":70},"01M6G9K2WJDDFPRYJ2668A9NMP","openhands",null,"Delegate coding to OpenHands CLI (model-agnostic, LiteLLM).","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\u002F7a\u002F7abd61ab5c9be09f20c7f15caace5f7ee95722bc179fcca4bc4c730ea64dc598\u002FSKILL.md",[],[5,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33],"macos","cli","linux","codex","code","model","coding","hermes","hermes agent","litellm","delegate","opencode","Coding-Agent","claude-code","hermes-agent","2026-11-10 10:12:30",false,"---\nname: openhands\ndescription: Delegate coding to OpenHands CLI (model-agnostic, LiteLLM).\nversion: 0.1.0\nauthor: Tim Koepsel (xzessmedia), Hermes Agent\nlicense: MIT\nplatforms: [linux, macos]\nmetadata:\n  hermes:\n    tags: [Coding-Agent, OpenHands, Model-Agnostic, LiteLLM]\n    related_skills: [claude-code, codex, opencode, hermes-agent]\n---\n\n# OpenHands CLI\n\nDelegate coding tasks to the [OpenHands CLI](https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands) via the `terminal` tool. OpenHands is model-agnostic: any LiteLLM-supported provider (OpenAI, Anthropic, OpenRouter, DeepSeek, Ollama, vLLM, etc.).\n\nThis skill is the headless-mode wrapper for batch \u002F one-shot delegation. The interactive textual UI is not used from Hermes.\n\n## When to Use\n\n- User wants a coding task delegated to OpenHands specifically.\n- User wants a coding agent that can run on a non-Anthropic \u002F non-OpenAI provider (DeepSeek, Qwen, Ollama, vLLM, Nous, etc.) — sibling skills `claude-code` and `codex` are tied to one vendor.\n- Multi-step file edits + shell commands inside a workspace.\n\nFor Claude-native, prefer `claude-code`. For OpenAI-native, prefer `codex`. For Hermes-native subagents, use `delegate_task`.\n\n## Prerequisites\n\n1. Install upstream (requires Python 3.12+ and `uv`):\n\n   ```\n   terminal(command=\"uv tool install openhands --python 3.12\")\n   ```\n\n   Verify: `openhands --version` (currently `OpenHands CLI 1.16.0` \u002F `SDK v1.21.0` at time of writing).\n\n2. Pick a model and set env vars for `--override-with-envs`:\n\n   ```\n   export LLM_MODEL=openrouter\u002Fopenai\u002Fgpt-4o-mini       # or any LiteLLM slug\n   export LLM_API_KEY=[REDACTED]\n   export LLM_BASE_URL=https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1     # omit for native OpenAI\n   ```\n\n   `LLM_MODEL` uses LiteLLM's full slug. When the provider is OpenRouter the slug is doubly-prefixed: `openrouter\u002F\u003Cvendor>\u002F\u003Cmodel>` (e.g. `openrouter\u002Fanthropic\u002Fclaude-sonnet-4.5`). For native Anthropic: `anthropic\u002Fclaude-sonnet-4-5`. For native OpenAI: `openai\u002Fgpt-4o-mini`.\n\n3. Suppress the startup banner so JSON output isn't preceded by ASCII art:\n\n   ```\n   export OPENHANDS_SUPPRESS_BANNER=1\n   ```\n\n## How to Run\n\nAlways invoke through the `terminal` tool. Always pass `--headless --json --override-with-envs --exit-without-confirmation` for automation.\n\n### One-shot task\n\n```\nterminal(\n  command=\"OPENHANDS_SUPPRESS_BANNER=1 LLM_MODEL=openrouter\u002Fopenai\u002Fgpt-4o-mini LLM_API_KEY=[REDACTED] LLM_BASE_URL=https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1 openhands --headless --json --override-with-envs --exit-without-confirmation -t 'Add error handling to all API calls in src\u002F'\",\n  workdir=\"\u002Fpath\u002Fto\u002Fproject\",\n  timeout=600\n)\n```\n\n### Background for long tasks\n\n```\nterminal(command=\"\u003Csame as above>\", workdir=\"\u002Fpath\u002Fto\u002Fproject\", background=true, notify_on_complete=true)\nprocess(action=\"poll\", session_id=\"\u003Cid>\")\nprocess(action=\"log\", session_id=\"\u003Cid>\")\n```\n\n### Resume a previous conversation\n\nOpenHands prints `Conversation ID: \u003C32-hex>` and a `Hint: openhands --resume \u003Cdashed-uuid>` line at the end of each run. Use the dashed form to resume:\n\n```\nterminal(\n  command=\"OPENHANDS_SUPPRESS_BANNER=1 LLM_MODEL=... openhands --headless --json --override-with-envs --exit-without-confirmation --resume \u003Cdashed-uuid> -t 'Now fix the bug you found'\",\n  workdir=\"\u002Fpath\u002Fto\u002Fproject\"\n)\n```\n\n## Real Flag List\n\nVerified against `openhands --help` (CLI 1.16.0). Anything not in this table is not a flag — pass it via env var or settings file.\n\n| Flag | Effect |\n|------|--------|\n| `--headless` | No UI, requires `-t` or `-f`. Auto-approves all actions (no `--llm-approve` in this mode). |\n| `--json` | JSONL event stream (requires `--headless`). |\n| `-t TEXT` | Task prompt. |\n| `-f PATH` | Read task from file. |\n| `--resume [ID]` | Resume conversation. No ID → list recent. |\n| `--last` | Resume most recent (with `--resume`). |\n| `--override-with-envs` | Apply `LLM_API_KEY` \u002F `LLM_BASE_URL` \u002F `LLM_MODEL` env vars. Without this, OpenHands uses `~\u002F.openhands\u002Fsettings.json` and ignores the env. |\n| `--exit-without-confirmation` | Don't show the \"are you sure\" exit dialog. |\n| `--always-approve` \u002F `--yolo` | Auto-approve every action (default in `--headless`). |\n| `--llm-approve` | LLM-based security gate (interactive only — does NOT work in headless). |\n| `--version` \u002F `-v` | Print version and exit. |\n\n**There is no `--model`, `--max-iterations`, `--workspace`, `--sandbox`, `--sandbox-type` flag.** Model is `LLM_MODEL`. Workspace is the `workdir` you pass to the `terminal` tool. Sandbox \u002F runtime is the `RUNTIME` and `SANDBOX_VOLUMES` env vars.\n\n## JSON Event Schema\n\nWith `--json --headless`, OpenHands emits JSONL — one JSON object per line, plus a handful of non-JSON status lines (`Initializing agent...`, `Agent is working`, `Agent finished`, the final summary box, `Goodbye!`, `Conversation ID:`, `Hint:`). Filter for lines starting with `{`.\n\nTop-level `kind` field discriminates events:\n\n- `MessageEvent` — user \u002F agent text turn. `source` is `user` or `agent`.\n- `ActionEvent` — agent picked a tool. Read `tool_name` (`file_editor`, `terminal`, `finish`) and `action.kind` (`FileEditorAction`, `TerminalAction`, `FinishAction`).\n- `ObservationEvent` — tool result. `observation.is_error` is the success flag. `source` is `environment`.\n- `FinishAction` inside an `ActionEvent` carries the agent's final message in `action.message`.\n\nThe cli prints all stderr from LiteLLM\u002FAuthlib first — see Pitfalls. Parse only stdout, line by line, ignoring lines that don't start with `{`.\n\n## Pitfalls\n\n- **LiteLLM warnings on every invocation.** The CLI prints `bedrock-runtime` and `sagemaker-runtime` warnings to stderr because `botocore` isn't installed. Plus an Authlib deprecation. These are noise, not failures. Pipe stderr to `\u002Fdev\u002Fnull` or filter it out before showing the user.\n- **Banner spam.** Without `OPENHANDS_SUPPRESS_BANNER=1`, every run starts with a multi-line `+--+` ASCII box advertising the SDK. Always export it.\n- **`--override-with-envs` is mandatory for automation.** Without it, OpenHands ignores `LLM_API_KEY` \u002F `LLM_BASE_URL` \u002F `LLM_MODEL` and falls back to `~\u002F.openhands\u002Fsettings.json`. On a fresh install this file doesn't exist and the CLI hangs waiting for first-run setup.\n- **Model slug is LiteLLM's, not the provider's.** `openrouter\u002Fopenai\u002Fgpt-4o-mini` works; `openai\u002Fgpt-4o-mini` while pointed at OpenRouter does not. `anthropic\u002Fclaude-sonnet-4-5` (hyphen) is native Anthropic; `openrouter\u002Fanthropic\u002Fclaude-sonnet-4.5` (dot) is via OpenRouter. Get it wrong → cryptic LiteLLM 400.\n- **`pip install openhands-ai` is the wrong package.** That's the legacy V0 SDK. The new CLI is `uv tool install openhands --python 3.12`. There is no maintained conda package.\n- **Resume ID format is fiddly.** The CLI ends with `Conversation ID: f46573d9cfdb45e492ca189bde40019b` (no dashes) and then a `Hint: openhands --resume f46573d9-cfdb-45e4-92ca-189bde40019b` (with dashes). Use the dashed form.\n- **Headless ignores `--llm-approve`.** If you pass it, you get an argparse error. Headless mode hardcodes always-approve.\n- **No Windows support upstream.** The OpenHands docs require WSL on Windows. This skill is gated `[linux, macos]` accordingly.\n- **`~\u002F.openhands\u002Fconversations\u002F\u003Cid>\u002F` accumulates.** Each run persists a trajectory. Clean it up if running batches.\n- **Heavy install (~200 packages).** Use `uv tool install` (isolated venv) to avoid dependency conflicts with the active project.\n\n## Verification\n\n```\nterminal(\n  command=\"OPENHANDS_SUPPRESS_BANNER=1 LLM_MODEL=openrouter\u002Fopenai\u002Fgpt-4o-mini LLM_API_KEY=[REDACTED] LLM_BASE_URL=https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1 openhands --headless --json --override-with-envs --exit-without-confirmation -t 'Print the string OPENHANDS_OK to stdout via the terminal tool.'\",\n  workdir=\"~\u002F.hermes\u002Fcache\u002Fscratch\",\n  timeout=120\n)\n```\n\nIf the JSONL stream ends with a `FinishAction` whose `action.message` mentions `OPENHANDS_OK`, the install is working.\n\n## Related\n\n- [OpenHands GitHub](https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands)\n- [OpenHands CLI command reference](https:\u002F\u002Fdocs.openhands.dev\u002Fopenhands\u002Fusage\u002Fcli\u002Fcommand-reference)\n- Sibling skills: `claude-code` (Anthropic-only), `codex` (OpenAI-only), `opencode` (multi-provider via OpenCode), `hermes-agent` (Hermes subagents via `delegate_task`).\n","# OpenHands CLI\n\nDelegate coding tasks to the [OpenHands CLI](https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands) via the `terminal` tool. OpenHands is model-agnostic: any LiteLLM-supported provider (OpenAI, Anthropic, OpenRouter, DeepSeek, Ollama, vLLM, etc.).\n\nThis skill is the headless-mode wrapper for batch \u002F one-shot delegation. The interactive textual UI is not used from Hermes.\n\n## When to Use\n\n- User wants a coding task delegated to OpenHands specifically.\n- User wants a coding agent that can run on a non-Anthropic \u002F non-OpenAI provider (DeepSeek, Qwen, Ollama, vLLM, Nous, etc.) — sibling skills `claude-code` and `codex` are tied to one vendor.\n- Multi-step file edits + shell commands inside a workspace.\n\nFor Claude-native, prefer `claude-code`. For OpenAI-native, prefer `codex`. For Hermes-native subagents, use `delegate_task`.\n\n## Prerequisites\n\n1. Install upstream (requires Python 3.12+ and `uv`):\n\n   ```\n   terminal(command=\"uv tool install openhands --python 3.12\")\n   ```\n\n   Verify: `openhands --version` (currently `OpenHands CLI 1.16.0` \u002F `SDK v1.21.0` at time of writing).\n\n2. Pick a model and set env vars for `--override-with-envs`:\n\n   ```\n   export LLM_MODEL=openrouter\u002Fopenai\u002Fgpt-4o-mini       # or any LiteLLM slug\n   export LLM_API_KEY=[REDACTED]\n   export LLM_BASE_URL=https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1     # omit for native OpenAI\n   ```\n\n   `LLM_MODEL` uses LiteLLM's full slug. When the provider is OpenRouter the slug is doubly-prefixed: `openrouter\u002F\u003Cvendor>\u002F\u003Cmodel>` (e.g. `openrouter\u002Fanthropic\u002Fclaude-sonnet-4.5`). For native Anthropic: `anthropic\u002Fclaude-sonnet-4-5`. For native OpenAI: `openai\u002Fgpt-4o-mini`.\n\n3. Suppress the startup banner so JSON output isn't preceded by ASCII art:\n\n   ```\n   export OPENHANDS_SUPPRESS_BANNER=1\n   ```\n\n## How to Run\n\nAlways invoke through the `terminal` tool. Always pass `--headless --json --override-with-envs --exit-without-confirmation` for automation.\n\n### One-shot task\n\n```\nterminal(\n  command=\"OPENHANDS_SUPPRESS_BANNER=1 LLM_MODEL=openrouter\u002Fopenai\u002Fgpt-4o-mini LLM_API_KEY=[REDACTED] LLM_BASE_URL=https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1 openhands --headless --json --override-with-envs --exit-without-confirmation -t 'Add error handling to all API calls in src\u002F'\",\n  workdir=\"\u002Fpath\u002Fto\u002Fproject\",\n  timeout=600\n)\n```\n\n### Background for long tasks\n\n```\nterminal(command=\"\u003Csame as above>\", workdir=\"\u002Fpath\u002Fto\u002Fproject\", background=true, notify_on_complete=true)\nprocess(action=\"poll\", session_id=\"\u003Cid>\")\nprocess(action=\"log\", session_id=\"\u003Cid>\")\n```\n\n### Resume a previous conversation\n\nOpenHands prints `Conversation ID: \u003C32-hex>` and a `Hint: openhands --resume \u003Cdashed-uuid>` line at the end of each run. Use the dashed form to resume:\n\n```\nterminal(\n  command=\"OPENHANDS_SUPPRESS_BANNER=1 LLM_MODEL=... openhands --headless --json --override-with-envs --exit-without-confirmation --resume \u003Cdashed-uuid> -t 'Now fix the bug you found'\",\n  workdir=\"\u002Fpath\u002Fto\u002Fproject\"\n)\n```\n\n## Real Flag List\n\nVerified against `openhands --help` (CLI 1.16.0). Anything not in this table is not a flag — pass it via env var or settings file.\n\n| Flag | Effect |\n|------|--------|\n| `--headless` | No UI, requires `-t` or `-f`. Auto-approves all actions (no `--llm-approve` in this mode). |\n| `--json` | JSONL event stream (requires `--headless`). |\n| `-t TEXT` | Task prompt. |\n| `-f PATH` | Read task from file. |\n| `--resume [ID]` | Resume conversation. No ID → list recent. |\n| `--last` | Resume most recent (with `--resume`). |\n| `--override-with-envs` | Apply `LLM_API_KEY` \u002F `LLM_BASE_URL` \u002F `LLM_MODEL` env vars. Without this, OpenHands uses `~\u002F.openhands\u002Fsettings.json` and ignores the env. |\n| `--exit-without-confirmation` | Don't show the \"are you sure\" exit dialog. |\n| `--always-approve` \u002F `--yolo` | Auto-approve every action (default in `--headless`). |\n| `--llm-approve` | LLM-based security gate (interactive only — does NOT work in headless). |\n| `--version` \u002F `-v` | Print version and exit. |\n\n**There is no `--model`, `--max-iterations`, `--workspace`, `--sandbox`, `--sandbox-type` flag.** Model is `LLM_MODEL`. Workspace is the `workdir` you pass to the `terminal` tool. Sandbox \u002F runtime is the `RUNTIME` and `SANDBOX_VOLUMES` env vars.\n\n## JSON Event Schema\n\nWith `--json --headless`, OpenHands emits JSONL — one JSON object per line, plus a handful of non-JSON status lines (`Initializing agent...`, `Agent is working`, `Agent finished`, the final summary box, `Goodbye!`, `Conversation ID:`, `Hint:`). Filter for lines starting with `{`.\n\nTop-level `kind` field discriminates events:\n\n- `MessageEvent` — user \u002F agent text turn. `source` is `user` or `agent`.\n- `ActionEvent` — agent picked a tool. Read `tool_name` (`file_editor`, `terminal`, `finish`) and `action.kind` (`FileEditorAction`, `TerminalAction`, `FinishAction`).\n- `ObservationEvent` — tool result. `observation.is_error` is the success flag. `source` is `environment`.\n- `FinishAction` inside an `ActionEvent` carries the agent's final message in `action.message`.\n\nThe cli prints all stderr from LiteLLM\u002FAuthlib first — see Pitfalls. Parse only stdout, line by line, ignoring lines that don't start with `{`.\n\n## Pitfalls\n\n- **LiteLLM warnings on every invocation.** The CLI prints `bedrock-runtime` and `sagemaker-runtime` warnings to stderr because `botocore` isn't installed. Plus an Authlib deprecation. These are noise, not failures. Pipe stderr to `\u002Fdev\u002Fnull` or filter it out before showing the user.\n- **Banner spam.** Without `OPENHANDS_SUPPRESS_BANNER=1`, every run starts with a multi-line `+--+` ASCII box advertising the SDK. Always export it.\n- **`--override-with-envs` is mandatory for automation.** Without it, OpenHands ignores `LLM_API_KEY` \u002F `LLM_BASE_URL` \u002F `LLM_MODEL` and falls back to `~\u002F.openhands\u002Fsettings.json`. On a fresh install this file doesn't exist and the CLI hangs waiting for first-run setup.\n- **Model slug is LiteLLM's, not the provider's.** `openrouter\u002Fopenai\u002Fgpt-4o-mini` works; `openai\u002Fgpt-4o-mini` while pointed at OpenRouter does not. `anthropic\u002Fclaude-sonnet-4-5` (hyphen) is native Anthropic; `openrouter\u002Fanthropic\u002Fclaude-sonnet-4.5` (dot) is via OpenRouter. Get it wrong → cryptic LiteLLM 400.\n- **`pip install openhands-ai` is the wrong package.** That's the legacy V0 SDK. The new CLI is `uv tool install openhands --python 3.12`. There is no maintained conda package.\n- **Resume ID format is fiddly.** The CLI ends with `Conversation ID: f46573d9cfdb45e492ca189bde40019b` (no dashes) and then a `Hint: openhands --resume f46573d9-cfdb-45e4-92ca-189bde40019b` (with dashes). Use the dashed form.\n- **Headless ignores `--llm-approve`.** If you pass it, you get an argparse error. Headless mode hardcodes always-approve.\n- **No Windows support upstream.** The OpenHands docs require WSL on Windows. This skill is gated `[linux, macos]` accordingly.\n- **`~\u002F.openhands\u002Fconversations\u002F\u003Cid>\u002F` accumulates.** Each run persists a trajectory. Clean it up if running batches.\n- **Heavy install (~200 packages).** Use `uv tool install` (isolated venv) to avoid dependency conflicts with the active project.\n\n## Verification\n\n```\nterminal(\n  command=\"OPENHANDS_SUPPRESS_BANNER=1 LLM_MODEL=openrouter\u002Fopenai\u002Fgpt-4o-mini LLM_API_KEY=[REDACTED] LLM_BASE_URL=https:\u002F\u002Fopenrouter.ai\u002Fapi\u002Fv1 openhands --headless --json --override-with-envs --exit-without-confirmation -t 'Print the string OPENHANDS_OK to stdout via the terminal tool.'\",\n  workdir=\"~\u002F.hermes\u002Fcache\u002Fscratch\",\n  timeout=120\n)\n```\n\nIf the JSONL stream ends with a `FinishAction` whose `action.message` mentions `OPENHANDS_OK`, the install is working.\n\n## Related\n\n- [OpenHands GitHub](https:\u002F\u002Fgithub.com\u002FAll-Hands-AI\u002FOpenHands)\n- [OpenHands CLI command reference](https:\u002F\u002Fdocs.openhands.dev\u002Fopenhands\u002Fusage\u002Fcli\u002Fcommand-reference)\n- Sibling skills: `claude-code` (Anthropic-only), `codex` (OpenAI-only), `opencode` (multi-provider via OpenCode), `hermes-agent` (Hermes subagents via `delegate_task`).\n",{"name":5,"author":39,"license":15,"version":40,"metadata":41,"platforms":48,"description":7},"Tim Koepsel (xzessmedia), Hermes Agent","0.1.0",{"hermes":42},{"tags":43,"related_skills":47},[31,44,45,46],"OpenHands","Model-Agnostic","LiteLLM",[32,22,30,33],[21,19],"main","0d1434f89f4bc6b07e9d272accd8c1013768820f","archive\u002Fskill-md-reviewed\u002Fsha256\u002F7a\u002F7abd61ab5c9be09f20c7f15caace5f7ee95722bc179fcca4bc4c730ea64dc598",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":58,"prompt":68,"zipAvailable":69},[59,62,65],{"tool":60,"command":61},"npx","npx skills add zapabob\u002FSkills --skill openhands",{"tool":63,"command":64},"bunx","bunx skills add zapabob\u002FSkills --skill openhands",{"tool":66,"command":67},"pnpm","pnpm dlx skills add zapabob\u002FSkills --skill openhands","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002F7a\u002F7abd61ab5c9be09f20c7f15caace5f7ee95722bc179fcca4bc4c730ea64dc598\u002FSKILL.md",true,[],[72,105,139,165,192,209],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":91,"publishedAt":34},"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",[78,83,86],{"categoryNo":79,"categoryName":80,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000010","后端与接口","SCENARIO","使用场景",{"categoryNo":84,"categoryName":85,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":87,"categoryName":88,"taxonomyType":89,"taxonomyTypeDesc":90},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[74,92,19,20,93,21,94,95,96,97,98,99,100,26,27,101,49,102,103,104],"openclaw","github","windows","official","search","https","social-media","api","Media","twitter","social","via","post",{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":122,"publishedAt":34},"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",[111,114,117,118,119],{"categoryNo":112,"categoryName":113,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":115,"categoryName":116,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":84,"categoryName":85,"taxonomyType":81,"taxonomyTypeDesc":82},{"categoryNo":87,"categoryName":88,"taxonomyType":89,"taxonomyTypeDesc":90},{"categoryNo":120,"categoryName":121,"taxonomyType":89,"taxonomyTypeDesc":90},"01JBSEEDCAT000000000000036","内容创作者",[123,124,125,126,19,127,21,128,94,129,130,131,23,132,133,134,135,136,137,138],"writing","writing plans","plans","design","review","workflow","documentation","tasks","development","planning","code-review","subagent","write","superpowers","paths","test-driven-development",{"skillNo":140,"name":141,"nameZh":6,"description":142,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":143,"categories":144,"keywords":148,"publishedAt":34},"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",[145],{"categoryNo":146,"categoryName":147,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000016","图像与多媒体",[141,19,21,149,150,151,152,153,154,155,24,156,157,158,159,160,161,162,163,164],"openai","speech","multimodal","Transformers","ASR","Audio","text","speech recognition","use","general","multilingual","translation","english","identification","processing","Audio Processing",{"skillNo":166,"name":167,"nameZh":6,"description":168,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":169,"categories":170,"keywords":174,"publishedAt":34},"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",[171],{"categoryNo":172,"categoryName":173,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000013","分析与可视化",[175,19,176,177,94,178,179,180,181,182,183,24,184,185,186,187,188,189,190,191],"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":193,"name":194,"nameZh":6,"description":195,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":196,"categories":197,"keywords":199,"publishedAt":34},"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",[198],{"categoryNo":84,"categoryName":85,"taxonomyType":81,"taxonomyTypeDesc":82},[194,19,200,93,21,201,202,203,204,205,206,26,27,207,208],"terminal","automation","http","monitoring","cron","json","rss","devops","with",{"skillNo":210,"name":211,"nameZh":6,"description":212,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":213,"categories":214,"keywords":217,"publishedAt":34},"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",[215,216],{"categoryNo":79,"categoryName":80,"taxonomyType":81,"taxonomyTypeDesc":82},{"categoryNo":87,"categoryName":88,"taxonomyType":89,"taxonomyTypeDesc":90},[218,219,220,21,221,149,222,223,224,152,225,99,226,227,228,229,230,231,232,233],"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"]