[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78THFA9MQFQ5YTX93RTHF5Z":3,"skill-same:01M78THFA9MQFQ5YTX93RTHF5Z":66},{"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":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":32,"install":33,"resources":45},"01M78THFA9MQFQ5YTX93RTHF5Z","cli-and-sessions",null,"Use the persistent browser-use\u002Fbu CLI for browser automation, named sessions, profiles, cloud connect, tunnels, config, doctor\u002Fsetup\u002Finit, and CLI failure recovery.","VectorSpaceLab\u002FAREX-Skill","VectorSpaceLab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F180637102?v=4","01M7538C5WPSJGWF7KHR4EFZ9F","01M7538C6R2GFP53NPXTCZNBRT",351,"MIT","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: cli-and-sessions\ndescription: \"Use the persistent browser-use\u002Fbu CLI for browser automation,\n  named sessions, profiles, cloud connect, tunnels, config, doctor\u002Fsetup\u002Finit,\n  and CLI failure recovery.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: MIT\n---\n\n# CLI and Sessions\n\nUse this sub-skill when the user wants terminal-first Browser Use automation: `browser-use` or `bu` commands, persistent browser sessions, local\u002Fremote Chrome connection, cloud browser connection from the CLI, template generation, config, setup\u002Fdoctor checks, tunnels, profile sync, or `browser-use python`.\n\n## Route First\n\n- Use this sub-skill for CLI commands, daemon\u002Fsession lifecycle, `browser-use cloud ...`, `browser-use tunnel`, `browser-use profile`, `browser-use config`, `browser-use doctor`, `browser-use setup`, `browser-use init`, and `browser-use python`.\n- Route Python `Agent(...)` workflows, `agent.run()`, history objects, callbacks, prompt design, and initial actions to `..\u002Fagent-programming\u002FSKILL.md`.\n- Route Python `Browser`, `BrowserSession`, `BrowserProfile`, CDP\u002Fprofile\u002Fsession constructor parameters, downloads, HAR\u002Fvideo, domain filters, and direct browser control to `..\u002Fbrowser-control\u002FSKILL.md`.\n- Route custom actions, default tool behavior inside an Agent, file tools, sensitive data, and tool validation to `..\u002Ftools-and-actions\u002FSKILL.md`.\n- Route LLM provider selection, `ChatBrowserUse`, model credentials for Agents, structured output, extraction LLMs, and cost tracking to `..\u002Fllm-and-output\u002FSKILL.md`.\n- Route production `@sandbox`, Python `Browser(use_cloud=True)`, MCP deployment, skills loading, telemetry, and broader cloud integrations to `..\u002Fproduction-integrations\u002FSKILL.md`.\n\n## Safe Defaults\n\n- Prefer `browser-use` in docs and scripts; `bu`, `browseruse`, and `browser` are aliases for the same fast CLI entrypoint.\n- Start with help-only validation before launching browsers: `browser-use --help`, `browser-use init --help`, `browser-use cloud v2 --help`, then `browser-use doctor` if environment checks are acceptable.\n- Use `--json` for coding-agent parsers when command output must be consumed programmatically.\n- Use named sessions for parallel work: `browser-use --session NAME ...`; names should contain only letters, digits, hyphens, and underscores.\n- Close sessions explicitly with `browser-use --session NAME close` or `browser-use close --all` to avoid stale daemons.\n- Do not paste real API keys into logs. For cloud CLI, use `browser-use cloud login \u003Ckey>` or `browser-use config set api_key \"$BROWSER_USE_API_KEY\"`; the CLI reads cloud API keys from its config file, not directly from `BROWSER_USE_API_KEY`.\n- Use `BROWSER_USE_HOME` only for isolation\u002Ftesting; otherwise the CLI stores state under the user’s Browser Use home.\n\n## Install and Health Checks\n\n```bash\nuv pip install \"browser-use[cli]\"\nbrowser-use install\nbrowser-use doctor\n```\n\n- `browser-use install` installs Chromium through Playwright and includes Linux system dependencies when needed.\n- `browser-use setup` creates CLI home\u002Fconfig state, offers Chromium\u002Fprofile-use\u002Fcloudflared setup, and prints config status.\n- `browser-use setup --yes` is suitable for non-interactive setup when the user approved dependency installation.\n- `browser-use doctor` checks package import, browser profile availability, network connectivity, `cloudflared`, and `profile-use`.\n- If only command availability needs verification, run `scripts\u002Fcli_smoke.sh --help-only` from this sub-skill.\n\n## Core Browser Commands\n\n```bash\nbrowser-use open https:\u002F\u002Fexample.com\nbrowser-use state\nbrowser-use click 5\nbrowser-use input 3 \"john@example.com\"\nbrowser-use keys \"Enter\"\nbrowser-use screenshot page.png\nbrowser-use close\n```\n\n- `open` auto-prefixes plain domains with `https:\u002F\u002F`.\n- `state` prints page\u002Fviewport\u002Fscroll details plus indexed interactive elements; rerun it after navigation or page mutation because indices can change.\n- `click` accepts either one element index or two pixel coordinates: `browser-use click 5` or `browser-use click 120 240`.\n- `input` clicks an indexed element and clear-then-types text; pass `\"\"` when you need to clear only.\n- `screenshot` writes bytes to a path or emits base64 when no path is provided; use `--full` for full-page capture.\n- See `references\u002Fcli-reference.md` for the command matrix.\n\n## Sessions and Browser Modes\n\n```bash\nbrowser-use --session work --headed open https:\u002F\u002Fexample.com\nbrowser-use --session work state\nbrowser-use sessions\nbrowser-use --session work close\n```\n\n- The first command for a session starts a background daemon; subsequent commands talk to that daemon over a socket\u002FTCP channel.\n- Each named session has separate PID\u002Fsocket\u002Ftoken\u002Fstate files and keeps the browser alive across CLI calls.\n- `--headed` starts a visible browser; omit it for headless Chromium.\n- `--profile \"Default\"` launches real Chrome with a local profile; bare `--profile` uses `Default`.\n- `browser-use connect` discovers a running local Chrome CDP endpoint; `--cdp-url http:\u002F\u002F...` or `--cdp-url ws:\u002F\u002F...` connects to a known CDP endpoint.\n- `--cdp-url` and `--profile` are mutually exclusive; cloud connect also conflicts with `--connect`, `--cdp-url`, and `--profile`.\n- See `references\u002Fsession-recipes.md` for multi-session, CDP, profile, tunnel, cloud, and `python` recipes.\n\n## Tabs, Cookies, Waits, and Data\n\n```bash\nbrowser-use tab list\nbrowser-use tab new https:\u002F\u002Fexample.org\nbrowser-use tab switch 1\nbrowser-use cookies export cookies.json\nbrowser-use wait selector \"h1\" --timeout 5000\nbrowser-use get html --selector \"main\"\nbrowser-use eval \"document.title\"\n```\n\n- Use `tab` subcommands to manage pages in the current session without starting another browser.\n- Use `cookies get\u002Fset\u002Fclear\u002Fexport\u002Fimport` for session cookie state; treat exported cookie files as secrets.\n- Use `wait selector` or `wait text` before interacting with pages that load asynchronously.\n- Use `get title\u002Fhtml\u002Ftext\u002Fvalue\u002Fattributes\u002Fbbox` for deterministic data retrieval from the current page.\n- Use `eval` for JavaScript snippets; avoid evaluating untrusted code.\n- `extract` exists in the CLI parser but currently returns an implementation error, so route extraction workflows to Agent\u002FPython guidance instead.\n\n## Persistent Python Command\n\n```bash\nbrowser-use python \"x = 42\"\nbrowser-use python \"print(x)\"\nbrowser-use python \"print(browser.url)\"\nbrowser-use python --vars\nbrowser-use python --reset\nbrowser-use python --file automate.py\n```\n\n- The CLI Python namespace persists per browser session and includes `json`, `re`, `os`, `Path`, `asyncio`, and a synchronous `browser` wrapper.\n- The `browser` wrapper supports common browser actions such as `goto`, `click`, `type`, `input`, `upload`, `scroll`, `screenshot`, `keys`, and page properties like `url`, `title`, and `html`.\n- Use `browser-use python` for quick session automation; use `..\u002Fagent-programming\u002FSKILL.md` when the user needs an LLM-driven `Agent` script.\n- `--file` reads and executes a local Python file in the persistent namespace; review file contents before running.\n\n## Cloud CLI, Profiles, and Tunnels\n\n```bash\nbrowser-use cloud login \u003Capi-key>\nbrowser-use cloud connect\nbrowser-use state\nbrowser-use cloud v2 GET \u002Fbrowsers\nbrowser-use cloud v2 POST \u002Ftasks '{\"task\":\"Search for AI news\",\"url\":\"https:\u002F\u002Fgoogle.com\"}'\nbrowser-use cloud v2 poll \u003Ctask-id>\nbrowser-use close\n```\n\n- `cloud login` stores the API key in CLI config with restricted permissions where supported.\n- `cloud connect` provisions a cloud browser profile, starts the session daemon in cloud mode, and may print a live\u002FCDP URL.\n- `cloud v2` and `cloud v3` are generic REST passthroughs; 4xx responses exit with code `2` and may print an expected body suggestion from OpenAPI.\n- `cloud signup` flows exist for account creation\u002Fclaiming; do not run signup without explicit user approval.\n- `profile` delegates to the managed `profile-use` binary for local cookie\u002Fprofile sync; install or update it with `browser-use profile update` when doctor reports it missing.\n- `tunnel \u003Cport>` starts a Cloudflare quick tunnel for local apps; `tunnel list`, `tunnel stop \u003Cport>`, and `tunnel stop --all` manage tunnel processes.\n- For Python cloud browser parameters or `@sandbox`, route to `..\u002Fproduction-integrations\u002FSKILL.md` and `..\u002Fbrowser-control\u002FSKILL.md`.\n\n## Config and Templates\n\n```bash\nbrowser-use config list\nbrowser-use config set cloud_connect_proxy us\nbrowser-use config set cloud_connect_timeout 30\nbrowser-use config unset cloud_connect_timeout\nbrowser-use init --list\nbrowser-use init --template basic --output my_agent.py\n```\n\n- Valid config keys include `api_key`, `cloud_connect_profile_id`, `cloud_connect_proxy`, `cloud_connect_timeout`, and `cloud_connect_recording`.\n- `config set` validates known keys and coerces integer\u002Fboolean values based on the schema.\n- `init` fetches template metadata\u002Fcontent at runtime; offline failures usually mean network\u002Ftemplate service access is unavailable.\n- Use `--force` only when the user approved overwriting an existing generated file.\n\n## Troubleshooting Protocol\n\n1. Capture the exact command, exit code, stdout\u002Fstderr, OS, shell, and whether `browser-use` or `bu` was used.\n2. Run `browser-use --help` and the specific subcommand help when possible.\n3. Run `browser-use doctor` for install\u002Fbrowser\u002Fnetwork\u002Fprofile\u002Ftunnel checks if safe in the environment.\n4. Inspect active sessions with `browser-use sessions`; close the affected session and retry with a fresh named session.\n5. For config\u002Fcloud issues, run `browser-use config list` and verify `api_key` is set without printing secret values.\n6. For page element failures, rerun `browser-use state` and update indices; page changes invalidate earlier indices.\n7. For stale daemon\u002Fsocket issues, run `browser-use --session NAME close`, then retry; use `browser-use close --all` only when the user accepts closing every CLI browser session.\n8. Use `references\u002Ftroubleshooting.md` for symptom-specific fixes.\n\n## Verification Snippets\n\n```bash\n# Help-only smoke test; does not launch browsers or require credentials.\n.\u002Fscripts\u002Fcli_smoke.sh --help-only\n\n# Isolated CLI home for a temporary session.\nBROWSER_USE_HOME=\"$(mktemp -d)\" browser-use --session smoke --json sessions\n```\n\n- Do not run browser-launching commands in restricted CI unless Chromium\u002Fbrowser dependencies are available and the user approved environment mutation.\n- Do not run cloud commands that create browsers, tasks, profiles, or signup accounts without explicit user approval.\n","# CLI and Sessions\n\nUse this sub-skill when the user wants terminal-first Browser Use automation: `browser-use` or `bu` commands, persistent browser sessions, local\u002Fremote Chrome connection, cloud browser connection from the CLI, template generation, config, setup\u002Fdoctor checks, tunnels, profile sync, or `browser-use python`.\n\n## Route First\n\n- Use this sub-skill for CLI commands, daemon\u002Fsession lifecycle, `browser-use cloud ...`, `browser-use tunnel`, `browser-use profile`, `browser-use config`, `browser-use doctor`, `browser-use setup`, `browser-use init`, and `browser-use python`.\n- Route Python `Agent(...)` workflows, `agent.run()`, history objects, callbacks, prompt design, and initial actions to `..\u002Fagent-programming\u002FSKILL.md`.\n- Route Python `Browser`, `BrowserSession`, `BrowserProfile`, CDP\u002Fprofile\u002Fsession constructor parameters, downloads, HAR\u002Fvideo, domain filters, and direct browser control to `..\u002Fbrowser-control\u002FSKILL.md`.\n- Route custom actions, default tool behavior inside an Agent, file tools, sensitive data, and tool validation to `..\u002Ftools-and-actions\u002FSKILL.md`.\n- Route LLM provider selection, `ChatBrowserUse`, model credentials for Agents, structured output, extraction LLMs, and cost tracking to `..\u002Fllm-and-output\u002FSKILL.md`.\n- Route production `@sandbox`, Python `Browser(use_cloud=True)`, MCP deployment, skills loading, telemetry, and broader cloud integrations to `..\u002Fproduction-integrations\u002FSKILL.md`.\n\n## Safe Defaults\n\n- Prefer `browser-use` in docs and scripts; `bu`, `browseruse`, and `browser` are aliases for the same fast CLI entrypoint.\n- Start with help-only validation before launching browsers: `browser-use --help`, `browser-use init --help`, `browser-use cloud v2 --help`, then `browser-use doctor` if environment checks are acceptable.\n- Use `--json` for coding-agent parsers when command output must be consumed programmatically.\n- Use named sessions for parallel work: `browser-use --session NAME ...`; names should contain only letters, digits, hyphens, and underscores.\n- Close sessions explicitly with `browser-use --session NAME close` or `browser-use close --all` to avoid stale daemons.\n- Do not paste real API keys into logs. For cloud CLI, use `browser-use cloud login \u003Ckey>` or `browser-use config set api_key \"$BROWSER_USE_API_KEY\"`; the CLI reads cloud API keys from its config file, not directly from `BROWSER_USE_API_KEY`.\n- Use `BROWSER_USE_HOME` only for isolation\u002Ftesting; otherwise the CLI stores state under the user’s Browser Use home.\n\n## Install and Health Checks\n\n```bash\nuv pip install \"browser-use[cli]\"\nbrowser-use install\nbrowser-use doctor\n```\n\n- `browser-use install` installs Chromium through Playwright and includes Linux system dependencies when needed.\n- `browser-use setup` creates CLI home\u002Fconfig state, offers Chromium\u002Fprofile-use\u002Fcloudflared setup, and prints config status.\n- `browser-use setup --yes` is suitable for non-interactive setup when the user approved dependency installation.\n- `browser-use doctor` checks package import, browser profile availability, network connectivity, `cloudflared`, and `profile-use`.\n- If only command availability needs verification, run `scripts\u002Fcli_smoke.sh --help-only` from this sub-skill.\n\n## Core Browser Commands\n\n```bash\nbrowser-use open https:\u002F\u002Fexample.com\nbrowser-use state\nbrowser-use click 5\nbrowser-use input 3 \"john@example.com\"\nbrowser-use keys \"Enter\"\nbrowser-use screenshot page.png\nbrowser-use close\n```\n\n- `open` auto-prefixes plain domains with `https:\u002F\u002F`.\n- `state` prints page\u002Fviewport\u002Fscroll details plus indexed interactive elements; rerun it after navigation or page mutation because indices can change.\n- `click` accepts either one element index or two pixel coordinates: `browser-use click 5` or `browser-use click 120 240`.\n- `input` clicks an indexed element and clear-then-types text; pass `\"\"` when you need to clear only.\n- `screenshot` writes bytes to a path or emits base64 when no path is provided; use `--full` for full-page capture.\n- See `references\u002Fcli-reference.md` for the command matrix.\n\n## Sessions and Browser Modes\n\n```bash\nbrowser-use --session work --headed open https:\u002F\u002Fexample.com\nbrowser-use --session work state\nbrowser-use sessions\nbrowser-use --session work close\n```\n\n- The first command for a session starts a background daemon; subsequent commands talk to that daemon over a socket\u002FTCP channel.\n- Each named session has separate PID\u002Fsocket\u002Ftoken\u002Fstate files and keeps the browser alive across CLI calls.\n- `--headed` starts a visible browser; omit it for headless Chromium.\n- `--profile \"Default\"` launches real Chrome with a local profile; bare `--profile` uses `Default`.\n- `browser-use connect` discovers a running local Chrome CDP endpoint; `--cdp-url http:\u002F\u002F...` or `--cdp-url ws:\u002F\u002F...` connects to a known CDP endpoint.\n- `--cdp-url` and `--profile` are mutually exclusive; cloud connect also conflicts with `--connect`, `--cdp-url`, and `--profile`.\n- See `references\u002Fsession-recipes.md` for multi-session, CDP, profile, tunnel, cloud, and `python` recipes.\n\n## Tabs, Cookies, Waits, and Data\n\n```bash\nbrowser-use tab list\nbrowser-use tab new https:\u002F\u002Fexample.org\nbrowser-use tab switch 1\nbrowser-use cookies export cookies.json\nbrowser-use wait selector \"h1\" --timeout 5000\nbrowser-use get html --selector \"main\"\nbrowser-use eval \"document.title\"\n```\n\n- Use `tab` subcommands to manage pages in the current session without starting another browser.\n- Use `cookies get\u002Fset\u002Fclear\u002Fexport\u002Fimport` for session cookie state; treat exported cookie files as secrets.\n- Use `wait selector` or `wait text` before interacting with pages that load asynchronously.\n- Use `get title\u002Fhtml\u002Ftext\u002Fvalue\u002Fattributes\u002Fbbox` for deterministic data retrieval from the current page.\n- Use `eval` for JavaScript snippets; avoid evaluating untrusted code.\n- `extract` exists in the CLI parser but currently returns an implementation error, so route extraction workflows to Agent\u002FPython guidance instead.\n\n## Persistent Python Command\n\n```bash\nbrowser-use python \"x = 42\"\nbrowser-use python \"print(x)\"\nbrowser-use python \"print(browser.url)\"\nbrowser-use python --vars\nbrowser-use python --reset\nbrowser-use python --file automate.py\n```\n\n- The CLI Python namespace persists per browser session and includes `json`, `re`, `os`, `Path`, `asyncio`, and a synchronous `browser` wrapper.\n- The `browser` wrapper supports common browser actions such as `goto`, `click`, `type`, `input`, `upload`, `scroll`, `screenshot`, `keys`, and page properties like `url`, `title`, and `html`.\n- Use `browser-use python` for quick session automation; use `..\u002Fagent-programming\u002FSKILL.md` when the user needs an LLM-driven `Agent` script.\n- `--file` reads and executes a local Python file in the persistent namespace; review file contents before running.\n\n## Cloud CLI, Profiles, and Tunnels\n\n```bash\nbrowser-use cloud login \u003Capi-key>\nbrowser-use cloud connect\nbrowser-use state\nbrowser-use cloud v2 GET \u002Fbrowsers\nbrowser-use cloud v2 POST \u002Ftasks '{\"task\":\"Search for AI news\",\"url\":\"https:\u002F\u002Fgoogle.com\"}'\nbrowser-use cloud v2 poll \u003Ctask-id>\nbrowser-use close\n```\n\n- `cloud login` stores the API key in CLI config with restricted permissions where supported.\n- `cloud connect` provisions a cloud browser profile, starts the session daemon in cloud mode, and may print a live\u002FCDP URL.\n- `cloud v2` and `cloud v3` are generic REST passthroughs; 4xx responses exit with code `2` and may print an expected body suggestion from OpenAPI.\n- `cloud signup` flows exist for account creation\u002Fclaiming; do not run signup without explicit user approval.\n- `profile` delegates to the managed `profile-use` binary for local cookie\u002Fprofile sync; install or update it with `browser-use profile update` when doctor reports it missing.\n- `tunnel \u003Cport>` starts a Cloudflare quick tunnel for local apps; `tunnel list`, `tunnel stop \u003Cport>`, and `tunnel stop --all` manage tunnel processes.\n- For Python cloud browser parameters or `@sandbox`, route to `..\u002Fproduction-integrations\u002FSKILL.md` and `..\u002Fbrowser-control\u002FSKILL.md`.\n\n## Config and Templates\n\n```bash\nbrowser-use config list\nbrowser-use config set cloud_connect_proxy us\nbrowser-use config set cloud_connect_timeout 30\nbrowser-use config unset cloud_connect_timeout\nbrowser-use init --list\nbrowser-use init --template basic --output my_agent.py\n```\n\n- Valid config keys include `api_key`, `cloud_connect_profile_id`, `cloud_connect_proxy`, `cloud_connect_timeout`, and `cloud_connect_recording`.\n- `config set` validates known keys and coerces integer\u002Fboolean values based on the schema.\n- `init` fetches template metadata\u002Fcontent at runtime; offline failures usually mean network\u002Ftemplate service access is unavailable.\n- Use `--force` only when the user approved overwriting an existing generated file.\n\n## Troubleshooting Protocol\n\n1. Capture the exact command, exit code, stdout\u002Fstderr, OS, shell, and whether `browser-use` or `bu` was used.\n2. Run `browser-use --help` and the specific subcommand help when possible.\n3. Run `browser-use doctor` for install\u002Fbrowser\u002Fnetwork\u002Fprofile\u002Ftunnel checks if safe in the environment.\n4. Inspect active sessions with `browser-use sessions`; close the affected session and retry with a fresh named session.\n5. For config\u002Fcloud issues, run `browser-use config list` and verify `api_key` is set without printing secret values.\n6. For page element failures, rerun `browser-use state` and update indices; page changes invalidate earlier indices.\n7. For stale daemon\u002Fsocket issues, run `browser-use --session NAME close`, then retry; use `browser-use close --all` only when the user accepts closing every CLI browser session.\n8. Use `references\u002Ftroubleshooting.md` for symptom-specific fixes.\n\n## Verification Snippets\n\n```bash\n# Help-only smoke test; does not launch browsers or require credentials.\n.\u002Fscripts\u002Fcli_smoke.sh --help-only\n\n# Isolated CLI home for a temporary session.\nBROWSER_USE_HOME=\"$(mktemp -d)\" browser-use --session smoke --json sessions\n```\n\n- Do not run browser-launching commands in restricted CI unless Chromium\u002Fbrowser dependencies are available and the user approved environment mutation.\n- Do not run cloud commands that create browsers, tasks, profiles, or signup accounts without explicit user approval.\n",{"name":5,"license":14,"metadata":23,"description":7,"disable-model-invocation":25},{"disco-role":24},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions",27,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab","2026-09-03 10:25:21",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":25},[35,38,41],{"tool":36,"command":37},"npx","npx skills add VectorSpaceLab\u002FAREX-Skill --skill cli-and-sessions",{"tool":39,"command":40},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill cli-and-sessions",{"tool":42,"command":43},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill cli-and-sessions","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions\u002FSKILL.md",[46,52,56,60],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Fcli-reference.md","REFERENCE","参考文档",10031,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions\u002Freferences\u002Fcli-reference.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fsession-recipes.md",9059,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions\u002Freferences\u002Fsession-recipes.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Ftroubleshooting.md",10346,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions\u002Freferences\u002Ftroubleshooting.md",{"path":61,"resourceType":62,"resourceTypeDesc":63,"sizeBytes":64,"sourceHtmlUrl":65},"scripts\u002Fcli_smoke.sh","SCRIPT","脚本",2535,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fbrowser-use\u002Fsub-skills\u002Fcli-and-sessions\u002Fscripts\u002Fcli_smoke.sh",[67,75,82,89,96,103],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":71,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":18},"01M799NYSH1R3RVAHSM2D88Y84","passnet-triton-opt","Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block\u002Fgrid\u002Fwarp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.","Apache-2.0","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-triton-opt\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":71,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":18},"01M799NS5APEC0N1FZ7XZVW6BN","passnet-solve","END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-solve\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":71,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":18},"01M799NNY9BKSM21V5R7QVTBJX","passnet-skill","PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under .\u002Fpass_dir\u002F, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-skill\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":71,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":18},"01M799NHH6NP7Z1RK77KGXAN6S","passnet-pattern-fusion","Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()\u002Freplacement_args()\u002Freplacement_func() files, when a pass \"failed to match\", or when deciding how to fuse multiple kernels\u002Fops.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-pattern-fusion\u002FSKILL.md",[],[],{"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":71,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":18},"01M799ND2J4WNC2PGAZ0A06NDF","passnet-orchestrate","Round\u002Fbatch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep\u002Frevert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-orchestrate\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":71,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":18},"01M799N8SBTMY02ZDG2AZF5ZWE","passnet-feedback","Fast, deterministic feedback for PassNet work: pre-flight pattern\u002Fmatch verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-feedback\u002FSKILL.md",[],[]]