[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CSYX08AP364D0PZ4E2SZ5H":3,"skill-same:01M7CSYX08AP364D0PZ4E2SZ5H":169},{"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":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":28,"commitSha":29,"skillDirectory":30,"forksCount":31,"repositoryHtmlUrl":32,"repositoryHomepageUrl":33,"ownerHtmlUrl":34,"ownerBlogUrl":6,"ownerDescription":35,"updatedAt":36,"install":37,"resources":50},"01M7CSYX08AP364D0PZ4E2SZ5H","vss-deploy-profile",null,"Use to select, configure, deploy, verify, debug, or tear down a VSS profile (base, search, lvs, warehouse, edge). Not for standalone microservices — use the vss-deploy-* skill.","YYC-Cube\u002FYYC3-Knowledge-Base","YYC-Cube","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F235006543?v=4","01M79J2PH4SWD2X0825X6PN748","yyc3","01M79J2PHVK3F6ZBYT8K2SCZCW",1,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: vss-deploy-profile\ndescription: Use to select, configure, deploy, verify, debug, or tear down a VSS profile (base, search, lvs, warehouse, edge). Not for standalone microservices — use the vss-deploy-* skill.\nlicense: Apache-2.0\nmetadata:\n  version: \"3.2.0\"\n  github-url: \"https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Fvideo-search-and-summarization\"\n  tags: \"nvidia blueprint deployment\"\n---\n# VSS Deploy\n\n## Purpose\n\nDeploy any VSS profile (`base`, `search`, `lvs`, `warehouse`, `alerts`, `edge`) using a compose-centric workflow: build env overrides, generate resolved compose (dry-run), review, then deploy. This SKILL.md covers the cross-profile concerns (**profile routing**, **prerequisites**, **NGC**, **GPU setup**, and the deploy\u002Fteardown flow). Profile-specific service lists, sizing, env recipes, endpoints, and debugging live in per-profile reference docs — load the one that matches the user's intent.\n\nHelper script: `run_script(\"scripts\u002Fnormalize_resolved_yml.py\", \"\u003Cresolved.yml>\")` normalizes a `docker compose config` dry-run dump for diff-friendly review during Step 3c. All other deployment work goes through `compose` \u002F `dev-profile.sh`.\n\n## Available Scripts\n\n| Script | Purpose | Arguments |\n|---|---|---|\n| `scripts\u002Fnormalize_resolved_yml.py` | Strip optional `depends_on` entries for services filtered out of `resolved.yml` before deploy. | Path to `resolved.yml` |\n\n## Profile Routing\n\nMatch the user's request to a profile, then load that profile's reference for sizing, services, env recipes, and debugging.\n\n| User says | Profile | Reference |\n|---|---|---|\n| \"deploy vss\" \u002F \"deploy base\" | `base` | [`references\u002Fbase.md`](references\u002Fbase.md) |\n| \"deploy alerts\" \u002F \"alert verification\" \u002F \"real-time alerts\" \u002F \"deploy for incident report\" | `alerts` | [`references\u002Falerts.md`](references\u002Falerts.md) |\n| \"deploy lvs\" \u002F \"video summarization\" | `lvs` | [`references\u002Flvs-profile.md`](references\u002Flvs-profile.md) |\n| \"deploy search\" \u002F \"video search\" | `search` | [`references\u002Fsearch.md`](references\u002Fsearch.md) |\n| \"deploy warehouse\" \u002F \"warehouse blueprint\" \u002F \"vss warehouse\" | `warehouse` | [`references\u002Fwarehouse.md`](references\u002Fwarehouse.md) |\n| \"debug warehouse\" \u002F \"warehouse not working\" \u002F \"warehouse FPS low\" \u002F \"warehouse BEV out of sync\" | `warehouse` (debug) | [`references\u002Fwarehouse-debug.md`](references\u002Fwarehouse-debug.md) |\n\n**Edge hardware routing** (DGX Spark, AGX\u002FIGX Thor): see [`references\u002Fedge.md`](references\u002Fedge.md). DGX Spark uses the Spark Nano 9B standalone local LLM on port `30081`; AGX\u002FIGX Thor uses the Edge 4B standalone vLLM fallback.\n\n**Each profile's reference owns its sizing table.** Don't pick a deployment shape from this file — open the profile reference and check minimum GPU count for the host's hardware against the (mode × platform) matrix there.\n\n\n## Instructions\n\nThe deployment flow is always: copy `.env` to `generated.env`, apply overrides, dry-run compose into `resolved.yml`, review, normalize, deploy, then wait for readiness.\n\n```bash\n# 1. cp dev-profile-\u003Cprofile>\u002F.env dev-profile-\u003Cprofile>\u002Fgenerated.env  (clean copy)\n# 2. Apply env overrides to generated.env  (source .env stays untouched)\n# 3. docker compose --env-file generated.env config > resolved.yml      (dry-run)\n# 4. Review resolved.yml\n# 5. docker compose --env-file generated.env -f resolved.yml up -d\n```\n\nThe source `.env` is treated as **read-only defaults** committed to the repo. The skill's per-deploy working copy is `generated.env` — same pattern `dev-profile.sh` uses internally. This keeps the checked-in `.env` clean across iterations.\n\n## Prerequisites\n\n1. **Repo path** — find `video-search-and-summarization\u002F` on disk. Check `TOOLS.md` if available.\n2. **NGC CLI & API key** — see [`references\u002Fngc.md`](references\u002Fngc.md). Confirm `$NGC_CLI_API_KEY` is set.\n3. **System prerequisites (GPU driver, Docker, NVIDIA Container Toolkit, kernel sysctls)** — full checks in [`references\u002Fprerequisites.md`](references\u002Fprerequisites.md). Canonical hardware\u002Fdriver matrix is the [VSS prerequisites page](https:\u002F\u002Fdocs.nvidia.com\u002Fvss\u002F3.2.0\u002Fprerequisites.html).\n\n### Pre-flight check\n\nRun before every deploy. The full system checklist and remediation steps live\nin [`references\u002Fprerequisites.md`](references\u002Fprerequisites.md#preflight).\nFor DGX Spark \u002F IGX Thor \u002F AGX Thor, also run the cache-cleaner check in\n[`references\u002Fedge.md`](references\u002Fedge.md#cache-cleaner-every-edge-deploy).\n\n**Detect sudo mode first.** Several pre-flight remediations and the\nedge cache-cleaner installer call `sudo`. If the host requires a\nsudo password, those steps will silently no-op under `sudo -n` and\nleave the deploy in a half-prepared state.\n\n```bash\nif sudo -n true 2>\u002Fdev\u002Fnull; then\n  echo \"passwordless sudo — pre-flight will auto-install missing pieces\"\nelse\n  echo \"sudo requires password — pre-flight will NOT auto-install; hand commands to the user\"\nfi\n```\n\nWhen sudo needs a password, the skill **must not** run privileged\ninstallers itself. Surface the copy-pasteable command block from\n`references\u002Fprerequisites.md` to the user with a *\"run this once and\nconfirm\"* handoff, then resume after the user replies.\n\nMinimum smoke test (must succeed):\n\n```bash\nnvidia-smi --query-gpu=index,name --format=csv,noheader\ndocker info 2>\u002Fdev\u002Fnull | grep -qi runtimes \\\n  && docker run --rm --gpus all ubuntu:22.04 nvidia-smi >\u002Fdev\u002Fnull 2>&1 \\\n  && echo \"nvidia runtime OK\"\n```\n\nIf the smoke test fails, do not proceed; open\n[`references\u002Fprerequisites.md`](references\u002Fprerequisites.md#preflight)\nfor the remediation tree.\n\n## Model Selection\n\n- `$LLM_REMOTE_URL` \u002F `$VLM_REMOTE_URL` if the user asks for remote\n- `$NGC_CLI_API_KEY` (local NIMs) or `$NVIDIA_API_KEY` (remote)\n\nIf no combination on this host satisfies the profile's sizing requirements, **stop and report the blocker** — don't silently pick another shape.\n\n> **Edge shared mode is platform-specific.** On DGX Spark, run `nvcr.io\u002Fnim\u002Fnvidia\u002Fnvidia-nemotron-nano-9b-v2-dgx-spark:1.0.0-variant` as a standalone local NIM on port `30081` and point the agent at it with `LLM_MODE=remote`. On AGX\u002FIGX Thor, keep using the Edge 4B standalone vLLM fallback with `HF_TOKEN`. Full recipes are in [`references\u002Fedge.md`](references\u002Fedge.md).\n\n## Deployment Flow\n\nAlways follow this sequence. Never skip the dry-run.\n\n### Step 0 — Tear down any existing deployment + clear data volumes\n\nIf a deployment already exists, tear it down AND clear stale data volumes before redeploying. \n\nFull procedure lives in [`references\u002Fteardown.md`](references\u002Fteardown.md).\n\n### Step 0a — Credentials gate (run before any env mutation)\n\nValidate every credential the chosen profile needs **before** Step 1c copies `.env` to `generated.env`. A 401 here is a 30-second failure; the same 401 inside a NIM cold-start is a 10–20 min failure. Run the discovery and probe flow in [`references\u002Fcredentials.md`](references\u002Fcredentials.md), then map the result against the chosen mode: missing or invalid required credentials are blockers, optional credentials are not.\n\n### Step 1 — Gather context\n\nBefore building env overrides, confirm:\n\n| Value | How to determine |\n|---|---|\n| **Profile** | Match user intent to the routing table above. Default: `base` |\n| **Repo path** | Find `video-search-and-summarization\u002F` on disk |\n| **Hardware** | `nvidia-smi --query-gpu=name,memory.total --format=csv,noheader` |\n| **LLM\u002FVLM placement** | Cross-reference available GPUs against the chosen profile's **Minimum GPU count** table |\n| **API keys** | `NGC_CLI_API_KEY` for local NIMs, `NVIDIA_API_KEY` for remote |\n| **`HOST_IP`** | `hostname -I \\| awk '{print $1}'` — the host's primary internal IP |\n| **`EXTERNAL_IP`** | Browser-reachable host\u002FIP. On Brev, use the secure-link domain (see [`references\u002Fbrev.md`](references\u002Fbrev.md)). |\n| **`HAPROXY_PORT`** | Browser-facing ingress port. Default `7777`; ensure it is free. |\n\nBefore `docker compose up`, verify `EXTERNAL_IP`, `HAPROXY_PORT`, `VSS_PUBLIC_HOST`, and `VSS_PUBLIC_PORT` are populated with browser-reachable values. Otherwise the stack may appear healthy while UI\u002FAPI\u002FVST links 404 or loop through Cloudflare Access.\n\n### Step 1b — Prepare the data directory\n\nLayout (asset paths, ownership, mount points, profile-specific subdirs) is documented in [`references\u002Fdata-directory.md`](references\u002Fdata-directory.md). Read that file before deploying for the first time on a host or when changing profiles.\n\n> **FORBIDDEN: `chown -R ubuntu:ubuntu $VSS_DATA_DIR` (or any recursive chown).**\n>\n> This is \"good housekeeping\" to a shell-admin instinct but is **the** deploy-breaking command in this stack. You will observe a \"healthy\" deploy (containers Up, endpoints 200) while the video pipeline is silently broken. Use `chmod -R 777` on the specific subdirs documented in `data-directory.md` — nothing else.\n\n### Step 1c — Initialize `generated.env`\n\nThe skill's per-deploy working copy. Always start from a fresh copy of the source `.env` — never mutate the source.\n\n```bash\nPROFILE=base\nENV_SRC=$REPO\u002Fdeploy\u002Fdocker\u002Fdeveloper-profiles\u002Fdev-profile-$PROFILE\u002F.env\nENV_GEN=$REPO\u002Fdeploy\u002Fdocker\u002Fdeveloper-profiles\u002Fdev-profile-$PROFILE\u002Fgenerated.env\n\ncp \"$ENV_SRC\" \"$ENV_GEN\"\n```\n\nAll subsequent writes (Brev `EXTERNAL_IP`, the env_overrides dict from Step 2) go to `$ENV_GEN`. `$ENV_SRC` is read-only from here on.\n\n### Step 1d — If deploying on Brev, set `EXTERNAL_IP` to the secure-link domain\n\nRead `BREV_ENV_ID` from `\u002Fetc\u002Fenvironment` and write `EXTERNAL_IP` into `generated.env` (NOT `.env`). Full secure-link behavior and troubleshooting are in [`references\u002Fbrev.md`](references\u002Fbrev.md).\n\n```bash\nbrev_env_id=$(awk -F= '\u002F^BREV_ENV_ID=\u002F {gsub(\u002F\"\u002F, \"\", $2); print $2; exit}' \u002Fetc\u002Fenvironment)\nsed -i \"s|^EXTERNAL_IP=.*|EXTERNAL_IP=7777-${brev_env_id}.brevlab.com|\" \"$ENV_GEN\"\n```\n\n### Step 2 — Build env_overrides\n\nProduce an `env_overrides` dict from the user request and the gathered context: choose remote\u002Flocal LLM\u002FVLM, set credentials, point at endpoints, set platform-specific flags. The full mapping (every override key, when it applies, defaults, profile-specific differences) lives in [`references\u002Fenv-overrides.md`](references\u002Fenv-overrides.md). Each profile reference has worked examples for that profile's common scenarios.\n\n### Step 3 — Apply overrides + dry-run\n\n**Working env file:** `\u003Crepo>\u002Fdeploy\u002Fdocker\u002Fdeveloper-profiles\u002Fdev-profile-\u003Cprofile>\u002Fgenerated.env` (created in Step 1c).\n\n> **Two env files, distinct roles.**\n> - `.env` — **read-only defaults**, checked in. Don't mutate it from the skill.\n> - `generated.env` — **the skill's per-deploy working copy**. All overrides (the dict from Step 2, plus the Brev `EXTERNAL_IP` from Step 1d) land here. `--env-file` always points at this file. Post-deploy verifiers should also read from `generated.env` for the actually-deployed values — see [Debugging a Deployment](#debugging-a-deployment).\n>\n> `generated.env` matches the convention `dev-profile.sh` uses internally — it's a per-invocation scratchpad regenerated by `cp .env generated.env` each run.\n\n```bash\n# (Step 1c already ran: cp $ENV_SRC $ENV_GEN)\n\n# Apply the env_overrides dict from Step 2 to generated.env\n# (read lines, update matching keys, append new keys, write)\n# Example:\n#   sed -i \"s|^LLM_MODE=.*|LLM_MODE=remote|\" \"$ENV_GEN\"\n#   sed -i \"s|^LLM_BASE_URL=.*|LLM_BASE_URL=http:\u002F\u002Flocalhost:30081|\" \"$ENV_GEN\"\n\n# Resolve compose\ncd $REPO\u002Fdeploy\u002Fdocker\ndocker compose --env-file $ENV_GEN config > resolved.yml\n```\n\nThe resolved YAML is saved to `\u003Crepo>\u002Fdeploy\u002Fdocker\u002Fresolved.yml`.\n\n### Step 3b — Verify resolved.yml has no unexpanded ${...} tokens\n\nUnexpanded `${VAR}` tokens in `resolved.yml` mean compose did not see those env values. Diagnostic procedure and common culprits live in [`references\u002Ftroubleshooting.md`](references\u002Ftroubleshooting.md).\n\n### Step 3c — Strip dangling optional `depends_on` from resolved.yml\n\n**MUST run after Step 3, before Step 5.** Skipping this aborts the deploy:\n\nNormalize - drop optional dependencies for services filtered out from resolved.yml\n\n```bash\n# From the repo root\nuv run skills\u002Fvss-deploy-profile\u002Fscripts\u002Fnormalize_resolved_yml.py \"$REPO\u002Fdeploy\u002Fdocker\u002Fresolved.yml\"\n```\nIf `uv` isn't on the host, install it once with `curl -LsSf https:\u002F\u002Fastral.sh\u002Fuv\u002Finstall.sh | sh` (no root needed).\n**Re-validate** before `up -d`:\n\n```bash\ndocker compose -f \"$REPO\u002Fdeploy\u002Fdocker\u002Fresolved.yml\" config --quiet && echo \"resolved.yml OK\"\n```\n\nIf validation still fails after the normalizer runs, capture the error and inspect — that's a different bug (a dependency that's not optional, or another schema violation), not the dangling-depends_on case.\n\n### Step 4 — Review\n\nShow the user a summary of what will be deployed:\n\n- Profile name and hardware\n- LLM\u002FVLM models and mode (local\u002Fremote\u002Flocal_shared)\n- Services that will start\n- GPU device assignment\n- Key endpoints (UI port, agent port)\n\nAsk: **\"Looks good — deploy now?\"** and wait for confirmation before Step 5.\n\n**Exception — autonomous mode.** If the user's request already asks you to run autonomously (e.g. \"deploy X autonomously\", \"run without confirmation\", \"non-interactive\"), skip the confirmation prompt and proceed straight to Step 5. This path exists so automated eval \u002F CI invocations don't hang waiting for a human reply they'll never get. In all other cases, a human must approve.\n\n### Step 5 — Deploy\n\n```bash\ncd $REPO\u002Fdeploy\u002Fdocker\ndocker compose --env-file $ENV_GEN -f resolved.yml up -d\n```\n\n> **`--env-file` is mandatory.** Without the same `generated.env` used in Step 3, `COMPOSE_PROFILES` may be unset and `up -d` can exit 0 with zero selected services.\n\n> **Do NOT use `--force-recreate` on retries.** It destroys already-warm NIM containers, forcing another 3–5 min torch.compile + CUDA-graph capture per NIM. If the previous `up -d` partially failed, fix the root cause (usually perms or an env typo) and just re-run `up -d` — Docker will re-create only the containers whose config changed or that are down.\n\n`docker compose up -d` only creates containers; it does not wait for internal services to finish warming. Never declare deploy success until the readiness gates pass.\n\n### Step 5b — Wait until the stack is actually healthy\n\n**Gate 0 — container count must be > 0.** Refuse to proceed past `up -d` until compose started the expected services:\n\n```bash\nexpected=$(docker compose --env-file $ENV_GEN -f resolved.yml config --services | wc -l)\nactual=$(docker compose -f resolved.yml ps -q | wc -l)\n[ \"$actual\" -gt 0 ] && [ \"$actual\" -ge \"$expected\" ] \\\n  || { echo \"FAIL: expected $expected services, got $actual — re-check Step 5 --env-file\"; exit 1; }\n```\n\nCold deploys can take 10–20 min. The full readiness procedure lives in [`references\u002Freadiness.md`](references\u002Freadiness.md), and each profile reference lists the required endpoints. **Never declare deploy done after `up -d`; only after every documented endpoint succeeds.**\n\n## Tear Down\n\n```bash\ncd $REPO\u002Fdeploy\u002Fdocker\ndocker compose -f resolved.yml down\n```\n\nFor switching profiles or recovering from a partial deploy, follow the full procedure in [`references\u002Fteardown.md`](references\u002Fteardown.md).\n\n## Debugging a Deployment\n\nUse this workflow when the user asks to \"debug the deploy\", \"verify it's working\", \"why is the agent not responding\", or similar. The goal is to confirm the full video-ingestion-to-agent-answer path, not just that containers are \"Up\".\n\nEach profile reference has a **Debugging** section listing the exact commands and failure-mode table for that profile.\n\n### Quick checks (all profiles)\n\n```bash\n# 1. All expected containers Up\ndocker ps --format 'table {{.Names}}\\t{{.Status}}'\n\n# 2. Agent API + UI responding\ncurl -sf http:\u002F\u002Flocalhost:8000\u002Fdocs >\u002Fdev\u002Fnull && echo \"agent OK\"\ncurl -sf http:\u002F\u002Flocalhost:3000\u002F >\u002Fdev\u002Fnull && echo \"ui OK\"\n\n# 3. VLM NIM responding (base\u002Flvs profiles)\ncurl -sf http:\u002F\u002Flocalhost:30082\u002Fv1\u002Fmodels | python3 -m json.tool\n\n# 4. LLM NIM responding\ncurl -sf http:\u002F\u002Flocalhost:30081\u002Fv1\u002Fmodels | python3 -m json.tool\n```\n\n### End-to-end video sanity check\n\nAfter the quick checks above pass, drive a real query through the agent — e.g. ask it over the REST API or UI to describe a video you've uploaded to VST. If the agent returns a non-empty answer, the upload → ingest → inference → reply path is healthy. If it fails, `docker logs vss-agent` shows which stage tripped.\n\n## Examples\n\n- Base profile, remote models: route to `base`, copy `dev-profile-base\u002F.env` to `generated.env`, set `LLM_MODE=remote` \u002F `VLM_MODE=remote`, dry-run, normalize, deploy, then verify `\u002Fdocs` and UI.\n- Search profile on RTX: route to `search`, follow [`references\u002Fsearch.md`](references\u002Fsearch.md) for sizing and endpoints, seed videos, then run the search-profile readiness checks.\n- Edge target: route through [`references\u002Fedge.md`](references\u002Fedge.md), then use the same `generated.env` → dry-run → normalize → deploy flow.\n\n## Limitations\n\n- This skill deploys compose-based VSS profiles only; standalone microservice deployment belongs to the matching `vss-deploy-*` skill.\n- Hardware sizing, model placement, and profile-specific readiness are owned by profile references; do not infer them from memory.\n- Privileged host remediation requires user approval when passwordless sudo is unavailable.\n\n## Troubleshooting\n\nStart with [`references\u002Fagent-failure-modes.md`](references\u002Fagent-failure-modes.md) for cross-profile failures such as NIM cold-start timeouts, OOM, remote endpoint 5xx responses, missing `NGC_CLI_API_KEY` \u002F `HF_TOKEN`, unexpanded values in `resolved.yml` etc.\n\n","# VSS Deploy\n\n## Purpose\n\nDeploy any VSS profile (`base`, `search`, `lvs`, `warehouse`, `alerts`, `edge`) using a compose-centric workflow: build env overrides, generate resolved compose (dry-run), review, then deploy. This SKILL.md covers the cross-profile concerns (**profile routing**, **prerequisites**, **NGC**, **GPU setup**, and the deploy\u002Fteardown flow). Profile-specific service lists, sizing, env recipes, endpoints, and debugging live in per-profile reference docs — load the one that matches the user's intent.\n\nHelper script: `run_script(\"scripts\u002Fnormalize_resolved_yml.py\", \"\u003Cresolved.yml>\")` normalizes a `docker compose config` dry-run dump for diff-friendly review during Step 3c. All other deployment work goes through `compose` \u002F `dev-profile.sh`.\n\n## Available Scripts\n\n| Script | Purpose | Arguments |\n|---|---|---|\n| `scripts\u002Fnormalize_resolved_yml.py` | Strip optional `depends_on` entries for services filtered out of `resolved.yml` before deploy. | Path to `resolved.yml` |\n\n## Profile Routing\n\nMatch the user's request to a profile, then load that profile's reference for sizing, services, env recipes, and debugging.\n\n| User says | Profile | Reference |\n|---|---|---|\n| \"deploy vss\" \u002F \"deploy base\" | `base` | [`references\u002Fbase.md`](references\u002Fbase.md) |\n| \"deploy alerts\" \u002F \"alert verification\" \u002F \"real-time alerts\" \u002F \"deploy for incident report\" | `alerts` | [`references\u002Falerts.md`](references\u002Falerts.md) |\n| \"deploy lvs\" \u002F \"video summarization\" | `lvs` | [`references\u002Flvs-profile.md`](references\u002Flvs-profile.md) |\n| \"deploy search\" \u002F \"video search\" | `search` | [`references\u002Fsearch.md`](references\u002Fsearch.md) |\n| \"deploy warehouse\" \u002F \"warehouse blueprint\" \u002F \"vss warehouse\" | `warehouse` | [`references\u002Fwarehouse.md`](references\u002Fwarehouse.md) |\n| \"debug warehouse\" \u002F \"warehouse not working\" \u002F \"warehouse FPS low\" \u002F \"warehouse BEV out of sync\" | `warehouse` (debug) | [`references\u002Fwarehouse-debug.md`](references\u002Fwarehouse-debug.md) |\n\n**Edge hardware routing** (DGX Spark, AGX\u002FIGX Thor): see [`references\u002Fedge.md`](references\u002Fedge.md). DGX Spark uses the Spark Nano 9B standalone local LLM on port `30081`; AGX\u002FIGX Thor uses the Edge 4B standalone vLLM fallback.\n\n**Each profile's reference owns its sizing table.** Don't pick a deployment shape from this file — open the profile reference and check minimum GPU count for the host's hardware against the (mode × platform) matrix there.\n\n\n## Instructions\n\nThe deployment flow is always: copy `.env` to `generated.env`, apply overrides, dry-run compose into `resolved.yml`, review, normalize, deploy, then wait for readiness.\n\n```bash\n# 1. cp dev-profile-\u003Cprofile>\u002F.env dev-profile-\u003Cprofile>\u002Fgenerated.env  (clean copy)\n# 2. Apply env overrides to generated.env  (source .env stays untouched)\n# 3. docker compose --env-file generated.env config > resolved.yml      (dry-run)\n# 4. Review resolved.yml\n# 5. docker compose --env-file generated.env -f resolved.yml up -d\n```\n\nThe source `.env` is treated as **read-only defaults** committed to the repo. The skill's per-deploy working copy is `generated.env` — same pattern `dev-profile.sh` uses internally. This keeps the checked-in `.env` clean across iterations.\n\n## Prerequisites\n\n1. **Repo path** — find `video-search-and-summarization\u002F` on disk. Check `TOOLS.md` if available.\n2. **NGC CLI & API key** — see [`references\u002Fngc.md`](references\u002Fngc.md). Confirm `$NGC_CLI_API_KEY` is set.\n3. **System prerequisites (GPU driver, Docker, NVIDIA Container Toolkit, kernel sysctls)** — full checks in [`references\u002Fprerequisites.md`](references\u002Fprerequisites.md). Canonical hardware\u002Fdriver matrix is the [VSS prerequisites page](https:\u002F\u002Fdocs.nvidia.com\u002Fvss\u002F3.2.0\u002Fprerequisites.html).\n\n### Pre-flight check\n\nRun before every deploy. The full system checklist and remediation steps live\nin [`references\u002Fprerequisites.md`](references\u002Fprerequisites.md#preflight).\nFor DGX Spark \u002F IGX Thor \u002F AGX Thor, also run the cache-cleaner check in\n[`references\u002Fedge.md`](references\u002Fedge.md#cache-cleaner-every-edge-deploy).\n\n**Detect sudo mode first.** Several pre-flight remediations and the\nedge cache-cleaner installer call `sudo`. If the host requires a\nsudo password, those steps will silently no-op under `sudo -n` and\nleave the deploy in a half-prepared state.\n\n```bash\nif sudo -n true 2>\u002Fdev\u002Fnull; then\n  echo \"passwordless sudo — pre-flight will auto-install missing pieces\"\nelse\n  echo \"sudo requires password — pre-flight will NOT auto-install; hand commands to the user\"\nfi\n```\n\nWhen sudo needs a password, the skill **must not** run privileged\ninstallers itself. Surface the copy-pasteable command block from\n`references\u002Fprerequisites.md` to the user with a *\"run this once and\nconfirm\"* handoff, then resume after the user replies.\n\nMinimum smoke test (must succeed):\n\n```bash\nnvidia-smi --query-gpu=index,name --format=csv,noheader\ndocker info 2>\u002Fdev\u002Fnull | grep -qi runtimes \\\n  && docker run --rm --gpus all ubuntu:22.04 nvidia-smi >\u002Fdev\u002Fnull 2>&1 \\\n  && echo \"nvidia runtime OK\"\n```\n\nIf the smoke test fails, do not proceed; open\n[`references\u002Fprerequisites.md`](references\u002Fprerequisites.md#preflight)\nfor the remediation tree.\n\n## Model Selection\n\n- `$LLM_REMOTE_URL` \u002F `$VLM_REMOTE_URL` if the user asks for remote\n- `$NGC_CLI_API_KEY` (local NIMs) or `$NVIDIA_API_KEY` (remote)\n\nIf no combination on this host satisfies the profile's sizing requirements, **stop and report the blocker** — don't silently pick another shape.\n\n> **Edge shared mode is platform-specific.** On DGX Spark, run `nvcr.io\u002Fnim\u002Fnvidia\u002Fnvidia-nemotron-nano-9b-v2-dgx-spark:1.0.0-variant` as a standalone local NIM on port `30081` and point the agent at it with `LLM_MODE=remote`. On AGX\u002FIGX Thor, keep using the Edge 4B standalone vLLM fallback with `HF_TOKEN`. Full recipes are in [`references\u002Fedge.md`](references\u002Fedge.md).\n\n## Deployment Flow\n\nAlways follow this sequence. Never skip the dry-run.\n\n### Step 0 — Tear down any existing deployment + clear data volumes\n\nIf a deployment already exists, tear it down AND clear stale data volumes before redeploying. \n\nFull procedure lives in [`references\u002Fteardown.md`](references\u002Fteardown.md).\n\n### Step 0a — Credentials gate (run before any env mutation)\n\nValidate every credential the chosen profile needs **before** Step 1c copies `.env` to `generated.env`. A 401 here is a 30-second failure; the same 401 inside a NIM cold-start is a 10–20 min failure. Run the discovery and probe flow in [`references\u002Fcredentials.md`](references\u002Fcredentials.md), then map the result against the chosen mode: missing or invalid required credentials are blockers, optional credentials are not.\n\n### Step 1 — Gather context\n\nBefore building env overrides, confirm:\n\n| Value | How to determine |\n|---|---|\n| **Profile** | Match user intent to the routing table above. Default: `base` |\n| **Repo path** | Find `video-search-and-summarization\u002F` on disk |\n| **Hardware** | `nvidia-smi --query-gpu=name,memory.total --format=csv,noheader` |\n| **LLM\u002FVLM placement** | Cross-reference available GPUs against the chosen profile's **Minimum GPU count** table |\n| **API keys** | `NGC_CLI_API_KEY` for local NIMs, `NVIDIA_API_KEY` for remote |\n| **`HOST_IP`** | `hostname -I \\| awk '{print $1}'` — the host's primary internal IP |\n| **`EXTERNAL_IP`** | Browser-reachable host\u002FIP. On Brev, use the secure-link domain (see [`references\u002Fbrev.md`](references\u002Fbrev.md)). |\n| **`HAPROXY_PORT`** | Browser-facing ingress port. Default `7777`; ensure it is free. |\n\nBefore `docker compose up`, verify `EXTERNAL_IP`, `HAPROXY_PORT`, `VSS_PUBLIC_HOST`, and `VSS_PUBLIC_PORT` are populated with browser-reachable values. Otherwise the stack may appear healthy while UI\u002FAPI\u002FVST links 404 or loop through Cloudflare Access.\n\n### Step 1b — Prepare the data directory\n\nLayout (asset paths, ownership, mount points, profile-specific subdirs) is documented in [`references\u002Fdata-directory.md`](references\u002Fdata-directory.md). Read that file before deploying for the first time on a host or when changing profiles.\n\n> **FORBIDDEN: `chown -R ubuntu:ubuntu $VSS_DATA_DIR` (or any recursive chown).**\n>\n> This is \"good housekeeping\" to a shell-admin instinct but is **the** deploy-breaking command in this stack. You will observe a \"healthy\" deploy (containers Up, endpoints 200) while the video pipeline is silently broken. Use `chmod -R 777` on the specific subdirs documented in `data-directory.md` — nothing else.\n\n### Step 1c — Initialize `generated.env`\n\nThe skill's per-deploy working copy. Always start from a fresh copy of the source `.env` — never mutate the source.\n\n```bash\nPROFILE=base\nENV_SRC=$REPO\u002Fdeploy\u002Fdocker\u002Fdeveloper-profiles\u002Fdev-profile-$PROFILE\u002F.env\nENV_GEN=$REPO\u002Fdeploy\u002Fdocker\u002Fdeveloper-profiles\u002Fdev-profile-$PROFILE\u002Fgenerated.env\n\ncp \"$ENV_SRC\" \"$ENV_GEN\"\n```\n\nAll subsequent writes (Brev `EXTERNAL_IP`, the env_overrides dict from Step 2) go to `$ENV_GEN`. `$ENV_SRC` is read-only from here on.\n\n### Step 1d — If deploying on Brev, set `EXTERNAL_IP` to the secure-link domain\n\nRead `BREV_ENV_ID` from `\u002Fetc\u002Fenvironment` and write `EXTERNAL_IP` into `generated.env` (NOT `.env`). Full secure-link behavior and troubleshooting are in [`references\u002Fbrev.md`](references\u002Fbrev.md).\n\n```bash\nbrev_env_id=$(awk -F= '\u002F^BREV_ENV_ID=\u002F {gsub(\u002F\"\u002F, \"\", $2); print $2; exit}' \u002Fetc\u002Fenvironment)\nsed -i \"s|^EXTERNAL_IP=.*|EXTERNAL_IP=7777-${brev_env_id}.brevlab.com|\" \"$ENV_GEN\"\n```\n\n### Step 2 — Build env_overrides\n\nProduce an `env_overrides` dict from the user request and the gathered context: choose remote\u002Flocal LLM\u002FVLM, set credentials, point at endpoints, set platform-specific flags. The full mapping (every override key, when it applies, defaults, profile-specific differences) lives in [`references\u002Fenv-overrides.md`](references\u002Fenv-overrides.md). Each profile reference has worked examples for that profile's common scenarios.\n\n### Step 3 — Apply overrides + dry-run\n\n**Working env file:** `\u003Crepo>\u002Fdeploy\u002Fdocker\u002Fdeveloper-profiles\u002Fdev-profile-\u003Cprofile>\u002Fgenerated.env` (created in Step 1c).\n\n> **Two env files, distinct roles.**\n> - `.env` — **read-only defaults**, checked in. Don't mutate it from the skill.\n> - `generated.env` — **the skill's per-deploy working copy**. All overrides (the dict from Step 2, plus the Brev `EXTERNAL_IP` from Step 1d) land here. `--env-file` always points at this file. Post-deploy verifiers should also read from `generated.env` for the actually-deployed values — see [Debugging a Deployment](#debugging-a-deployment).\n>\n> `generated.env` matches the convention `dev-profile.sh` uses internally — it's a per-invocation scratchpad regenerated by `cp .env generated.env` each run.\n\n```bash\n# (Step 1c already ran: cp $ENV_SRC $ENV_GEN)\n\n# Apply the env_overrides dict from Step 2 to generated.env\n# (read lines, update matching keys, append new keys, write)\n# Example:\n#   sed -i \"s|^LLM_MODE=.*|LLM_MODE=remote|\" \"$ENV_GEN\"\n#   sed -i \"s|^LLM_BASE_URL=.*|LLM_BASE_URL=http:\u002F\u002Flocalhost:30081|\" \"$ENV_GEN\"\n\n# Resolve compose\ncd $REPO\u002Fdeploy\u002Fdocker\ndocker compose --env-file $ENV_GEN config > resolved.yml\n```\n\nThe resolved YAML is saved to `\u003Crepo>\u002Fdeploy\u002Fdocker\u002Fresolved.yml`.\n\n### Step 3b — Verify resolved.yml has no unexpanded ${...} tokens\n\nUnexpanded `${VAR}` tokens in `resolved.yml` mean compose did not see those env values. Diagnostic procedure and common culprits live in [`references\u002Ftroubleshooting.md`](references\u002Ftroubleshooting.md).\n\n### Step 3c — Strip dangling optional `depends_on` from resolved.yml\n\n**MUST run after Step 3, before Step 5.** Skipping this aborts the deploy:\n\nNormalize - drop optional dependencies for services filtered out from resolved.yml\n\n```bash\n# From the repo root\nuv run skills\u002Fvss-deploy-profile\u002Fscripts\u002Fnormalize_resolved_yml.py \"$REPO\u002Fdeploy\u002Fdocker\u002Fresolved.yml\"\n```\nIf `uv` isn't on the host, install it once with `curl -LsSf https:\u002F\u002Fastral.sh\u002Fuv\u002Finstall.sh | sh` (no root needed).\n**Re-validate** before `up -d`:\n\n```bash\ndocker compose -f \"$REPO\u002Fdeploy\u002Fdocker\u002Fresolved.yml\" config --quiet && echo \"resolved.yml OK\"\n```\n\nIf validation still fails after the normalizer runs, capture the error and inspect — that's a different bug (a dependency that's not optional, or another schema violation), not the dangling-depends_on case.\n\n### Step 4 — Review\n\nShow the user a summary of what will be deployed:\n\n- Profile name and hardware\n- LLM\u002FVLM models and mode (local\u002Fremote\u002Flocal_shared)\n- Services that will start\n- GPU device assignment\n- Key endpoints (UI port, agent port)\n\nAsk: **\"Looks good — deploy now?\"** and wait for confirmation before Step 5.\n\n**Exception — autonomous mode.** If the user's request already asks you to run autonomously (e.g. \"deploy X autonomously\", \"run without confirmation\", \"non-interactive\"), skip the confirmation prompt and proceed straight to Step 5. This path exists so automated eval \u002F CI invocations don't hang waiting for a human reply they'll never get. In all other cases, a human must approve.\n\n### Step 5 — Deploy\n\n```bash\ncd $REPO\u002Fdeploy\u002Fdocker\ndocker compose --env-file $ENV_GEN -f resolved.yml up -d\n```\n\n> **`--env-file` is mandatory.** Without the same `generated.env` used in Step 3, `COMPOSE_PROFILES` may be unset and `up -d` can exit 0 with zero selected services.\n\n> **Do NOT use `--force-recreate` on retries.** It destroys already-warm NIM containers, forcing another 3–5 min torch.compile + CUDA-graph capture per NIM. If the previous `up -d` partially failed, fix the root cause (usually perms or an env typo) and just re-run `up -d` — Docker will re-create only the containers whose config changed or that are down.\n\n`docker compose up -d` only creates containers; it does not wait for internal services to finish warming. Never declare deploy success until the readiness gates pass.\n\n### Step 5b — Wait until the stack is actually healthy\n\n**Gate 0 — container count must be > 0.** Refuse to proceed past `up -d` until compose started the expected services:\n\n```bash\nexpected=$(docker compose --env-file $ENV_GEN -f resolved.yml config --services | wc -l)\nactual=$(docker compose -f resolved.yml ps -q | wc -l)\n[ \"$actual\" -gt 0 ] && [ \"$actual\" -ge \"$expected\" ] \\\n  || { echo \"FAIL: expected $expected services, got $actual — re-check Step 5 --env-file\"; exit 1; }\n```\n\nCold deploys can take 10–20 min. The full readiness procedure lives in [`references\u002Freadiness.md`](references\u002Freadiness.md), and each profile reference lists the required endpoints. **Never declare deploy done after `up -d`; only after every documented endpoint succeeds.**\n\n## Tear Down\n\n```bash\ncd $REPO\u002Fdeploy\u002Fdocker\ndocker compose -f resolved.yml down\n```\n\nFor switching profiles or recovering from a partial deploy, follow the full procedure in [`references\u002Fteardown.md`](references\u002Fteardown.md).\n\n## Debugging a Deployment\n\nUse this workflow when the user asks to \"debug the deploy\", \"verify it's working\", \"why is the agent not responding\", or similar. The goal is to confirm the full video-ingestion-to-agent-answer path, not just that containers are \"Up\".\n\nEach profile reference has a **Debugging** section listing the exact commands and failure-mode table for that profile.\n\n### Quick checks (all profiles)\n\n```bash\n# 1. All expected containers Up\ndocker ps --format 'table {{.Names}}\\t{{.Status}}'\n\n# 2. Agent API + UI responding\ncurl -sf http:\u002F\u002Flocalhost:8000\u002Fdocs >\u002Fdev\u002Fnull && echo \"agent OK\"\ncurl -sf http:\u002F\u002Flocalhost:3000\u002F >\u002Fdev\u002Fnull && echo \"ui OK\"\n\n# 3. VLM NIM responding (base\u002Flvs profiles)\ncurl -sf http:\u002F\u002Flocalhost:30082\u002Fv1\u002Fmodels | python3 -m json.tool\n\n# 4. LLM NIM responding\ncurl -sf http:\u002F\u002Flocalhost:30081\u002Fv1\u002Fmodels | python3 -m json.tool\n```\n\n### End-to-end video sanity check\n\nAfter the quick checks above pass, drive a real query through the agent — e.g. ask it over the REST API or UI to describe a video you've uploaded to VST. If the agent returns a non-empty answer, the upload → ingest → inference → reply path is healthy. If it fails, `docker logs vss-agent` shows which stage tripped.\n\n## Examples\n\n- Base profile, remote models: route to `base`, copy `dev-profile-base\u002F.env` to `generated.env`, set `LLM_MODE=remote` \u002F `VLM_MODE=remote`, dry-run, normalize, deploy, then verify `\u002Fdocs` and UI.\n- Search profile on RTX: route to `search`, follow [`references\u002Fsearch.md`](references\u002Fsearch.md) for sizing and endpoints, seed videos, then run the search-profile readiness checks.\n- Edge target: route through [`references\u002Fedge.md`](references\u002Fedge.md), then use the same `generated.env` → dry-run → normalize → deploy flow.\n\n## Limitations\n\n- This skill deploys compose-based VSS profiles only; standalone microservice deployment belongs to the matching `vss-deploy-*` skill.\n- Hardware sizing, model placement, and profile-specific readiness are owned by profile references; do not infer them from memory.\n- Privileged host remediation requires user approval when passwordless sudo is unavailable.\n\n## Troubleshooting\n\nStart with [`references\u002Fagent-failure-modes.md`](references\u002Fagent-failure-modes.md) for cross-profile failures such as NIM cold-start timeouts, OOM, remote endpoint 5xx responses, missing `NGC_CLI_API_KEY` \u002F `HF_TOKEN`, unexpanded values in `resolved.yml` etc.\n\n",{"name":5,"license":15,"metadata":24,"description":7},{"tags":25,"version":26,"github-url":27},"nvidia blueprint deployment","3.2.0","https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Fvideo-search-and-summarization","main","87afe6ce96220accf50a345dea46784c8c9f210e","YYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile",0,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base","https:\u002F\u002Fkb.yyc3.top\u002F","https:\u002F\u002Fgithub.com\u002FYYC-Cube","万象归元于云枢 | 深栈智启新纪元","2026-11-10 10:37:32",{"source":8,"skillName":5,"commands":38,"prompt":48,"zipAvailable":49},[39,42,45],{"tool":40,"command":41},"npx","npx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill vss-deploy-profile",{"tool":43,"command":44},"bunx","bunx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill vss-deploy-profile",{"tool":46,"command":47},"pnpm","pnpm dlx skills add YYC-Cube\u002FYYC3-Knowledge-Base --skill vss-deploy-profile","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002FSKILL.md",true,[51,57,61,65,69,73,77,81,85,89,93,97,101,105,109,113,117,121,125,129,133,137,141,145,149,153,159,163],{"path":52,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":55,"sourceHtmlUrl":56},"BENCHMARK.md","REFERENCE","参考文档",2645,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002FBENCHMARK.md",{"path":58,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":59,"sourceHtmlUrl":60},"evals\u002Falerts_cv.json",1771,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fevals\u002Falerts_cv.json",{"path":62,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":63,"sourceHtmlUrl":64},"evals\u002Falerts_vlm.json",1632,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fevals\u002Falerts_vlm.json",{"path":66,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":67,"sourceHtmlUrl":68},"evals\u002Fbase.json",1367,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fevals\u002Fbase.json",{"path":70,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":71,"sourceHtmlUrl":72},"evals\u002Flvs.json",1355,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fevals\u002Flvs.json",{"path":74,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":75,"sourceHtmlUrl":76},"evals\u002Fsearch.json",1385,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fevals\u002Fsearch.json",{"path":78,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":79,"sourceHtmlUrl":80},"evals\u002Fwarehouse.json",2696,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fevals\u002Fwarehouse.json",{"path":82,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":83,"sourceHtmlUrl":84},"references\u002Fagent-failure-modes.md",7068,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fagent-failure-modes.md",{"path":86,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":87,"sourceHtmlUrl":88},"references\u002Falerts.md",23756,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Falerts.md",{"path":90,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":91,"sourceHtmlUrl":92},"references\u002Fbase.md",28057,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fbase.md",{"path":94,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":95,"sourceHtmlUrl":96},"references\u002Fbrev.md",4915,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fbrev.md",{"path":98,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":99,"sourceHtmlUrl":100},"references\u002Fcredentials.md",2374,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fcredentials.md",{"path":102,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":103,"sourceHtmlUrl":104},"references\u002Fdata-directory.md",3224,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fdata-directory.md",{"path":106,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":107,"sourceHtmlUrl":108},"references\u002Fedge.md",11307,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fedge.md",{"path":110,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":111,"sourceHtmlUrl":112},"references\u002Fenv-overrides.md",4261,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fenv-overrides.md",{"path":114,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":115,"sourceHtmlUrl":116},"references\u002Fknowledge-retrieval.md",10102,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fknowledge-retrieval.md",{"path":118,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":119,"sourceHtmlUrl":120},"references\u002Flvs-profile.md",18008,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Flvs-profile.md",{"path":122,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":123,"sourceHtmlUrl":124},"references\u002Fngc.md",3038,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fngc.md",{"path":126,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":127,"sourceHtmlUrl":128},"references\u002Fprerequisites.md",10197,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fprerequisites.md",{"path":130,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":131,"sourceHtmlUrl":132},"references\u002Freadiness.md",2133,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Freadiness.md",{"path":134,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":135,"sourceHtmlUrl":136},"references\u002Fsearch.md",24722,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fsearch.md",{"path":138,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":139,"sourceHtmlUrl":140},"references\u002Fteardown.md",3075,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fteardown.md",{"path":142,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":143,"sourceHtmlUrl":144},"references\u002Ftroubleshooting.md",975,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Ftroubleshooting.md",{"path":146,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":147,"sourceHtmlUrl":148},"references\u002Fwarehouse-debug.md",29177,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fwarehouse-debug.md",{"path":150,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":151,"sourceHtmlUrl":152},"references\u002Fwarehouse.md",45746,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Freferences\u002Fwarehouse.md",{"path":154,"resourceType":155,"resourceTypeDesc":156,"sizeBytes":157,"sourceHtmlUrl":158},"scripts\u002Fnormalize_resolved_yml.py","SCRIPT","脚本",3990,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fscripts\u002Fnormalize_resolved_yml.py",{"path":160,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":161,"sourceHtmlUrl":162},"skill-card.md",3112,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fskill-card.md",{"path":164,"resourceType":165,"resourceTypeDesc":166,"sizeBytes":167,"sourceHtmlUrl":168},"skill.oms.sig","OTHER","其他",10229,"https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-Skills-技能包\u002FYYC3-Skills-技能包-skills\u002Fskills\u002Fvss-deploy-profile\u002Fskill.oms.sig",[170,178,185,192,199,206],{"skillNo":171,"name":172,"nameZh":6,"description":173,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":174,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":177,"publishedAt":19},"01M7CTDWHERCXPQCP0K03A86AS","vercel-sandbox","Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include \"Vercel Sandbox browser\", \"microVM Chrome\", \"agent-browser in sandbox\", \"browser automation on Vercel\", or any task requiring Chrome in a Vercel Sandbox.","MIT","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fvercel-sandbox\u002FSKILL.md",[],[],{"skillNo":179,"name":180,"nameZh":6,"description":181,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":174,"licenseFileUrl":6,"sourceHtmlUrl":182,"categories":183,"keywords":184,"publishedAt":19},"01M7CTD8D2BNJY54S65M47HEJA","slack","Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include \"check my Slack\", \"what channels have unreads\", \"send a message to\", \"search Slack for\", \"extract from Slack\", \"find who said\", or any task requiring programmatic Slack interaction.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fslack\u002FSKILL.md",[],[],{"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":174,"licenseFileUrl":6,"sourceHtmlUrl":189,"categories":190,"keywords":191,"publishedAt":19},"01M7CTCSSB8ZP3Q58C3YZDZK1W","electron","Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include \"automate Slack app\", \"control VS Code\", \"interact with Discord app\", \"test this Electron app\", \"connect to desktop app\", or any task requiring automation of a native Electron application.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Felectron\u002FSKILL.md",[],[],{"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":174,"licenseFileUrl":6,"sourceHtmlUrl":196,"categories":197,"keywords":198,"publishedAt":19},"01M7CTCBJY76688V4DCZGWFHGQ","dogfood","Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to \"dogfood\", \"QA\", \"exploratory test\", \"find issues\", \"bug hunt\", \"test this app\u002Fsite\u002Fplatform\", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fdogfood\u002FSKILL.md",[],[],{"skillNo":200,"name":201,"nameZh":6,"description":202,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":174,"licenseFileUrl":6,"sourceHtmlUrl":203,"categories":204,"keywords":205,"publishedAt":19},"01M7CTC4H0FFGVEAQR9PA4GA1A","agent-browser","Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to \"open a website\", \"fill out a form\", \"click a button\", \"take a screenshot\", \"scrape data from a page\", \"test this web app\", \"login to a site\", \"automate browser actions\", or any task requiring programmatic web interaction.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-WebSpec-Web标准\u002FYYC3-WebSpec-Web标准-standards\u002Fbrowser-agent\u002Fskills\u002Fagent-browser\u002FSKILL.md",[],[],{"skillNo":207,"name":208,"nameZh":6,"description":209,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":174,"licenseFileUrl":6,"sourceHtmlUrl":210,"categories":211,"keywords":212,"publishedAt":19},"01M7CTBKVFYBG78FDSABSDJTTQ","shadcn","Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn\u002Fui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for \"shadcn init\", \"create an app with --preset\", \"switch to --preset\", and for building a design system or applying a DESIGN.md or brand spec to shadcn\u002Fui.","https:\u002F\u002Fgithub.com\u002FYYC-Cube\u002FYYC3-Knowledge-Base\u002Fblob\u002Fmain\u002FYYC3-UI-界面组件\u002FYYC3-UI-界面组件-ui\u002Fskills\u002Fshadcn\u002FSKILL.md",[],[]]