deployment-ops
Guides CI/CD setup, deployment commands, observability, data ingestion, and Gemini Enterprise registration for Agent Starter Pack projects.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- deployment-ops
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Guides CI/CD setup, deployment commands, observability, data ingestion, and Gemini Enterprise registration for Agent Starter Pack projects.
- disable-model-invocation
- true
02 资源文件 7
- 参考文档references/ci-cd-reference.md1.8 KB
- 参考文档references/data-ingestion.md1.3 KB
- 参考文档references/gemini-enterprise.md1.3 KB
- 参考文档references/generated-project-commands.md1.6 KB
- 参考文档references/observability.md1.2 KB
- 参考文档references/troubleshooting.md1.7 KB
- 参考文档references/workflows.md2.1 KB
03 SKILL.md 正文
Deployment operations
Use this sub-skill when the user is turning an Agent Starter Pack project into a deployable, observable, or registered cloud asset.
Covers
setup-cicdand its Terraform/runner setup.- Generated-project deployment commands such as
make setup-dev-env,make setup-datastore,make deploy,make data-ingestion,make sync-data,make register-gemini-enterprise,make inspector, and the related local-development targets. - Data ingestion, observability, and Gemini Enterprise registration guidance.
- Cloud-provider prerequisites such as GitHub CLI, gcloud, Terraform, repository ownership, and generated deployment metadata.
Excludes
- New-project template selection belongs in
project-scaffolding. - In-place project edits and upgrades belong in
project-maintenance.
Read first
../../references/cli-reference.mdfor command routing.../../references/package-overview.mdfor install and sanity-check guidance.references/workflows.mdfor the end-to-end deployment flow.references/ci-cd-reference.mdforsetup-cicddetails.references/generated-project-commands.mdfor the generated Makefile command map.references/data-ingestion.mdfor RAG and datastore setup.references/observability.mdfor telemetry and BigQuery analytics.references/gemini-enterprise.mdfor registration flows.references/troubleshooting.mdfor cloud and auth failures.
Common workflow
- Confirm whether the user is asking about a generated project, not the source package itself.
- Decide whether the task is CI/CD bootstrap, runtime deployment, data ingestion, observability, or Gemini Enterprise registration.
- Check the relevant generated-project command or cloud prerequisite before suggesting a fix.
- Use the sub-reference pages to explain the concrete command sequence and likely failure points.
- Route project-generation questions back to
project-scaffoldingand project-maintenance questions back toproject-maintenance.
Useful signals
setup-cicd,Terraform,Cloud Build,GitHub Actions,gcloud,gh,local-statemake deploy,make setup-dev-env,make data-ingestion,make sync-data,make inspector,make register-gemini-enterpriseagent_engine,cloud_run,gke,deployment_metadata.json,agent card,Gemini Enterprisevertex_ai_search,vertex_ai_vector_search, telemetry, Cloud Trace, BigQuery analytics, observability
Validation mindset
- Do not claim cloud readiness from package importability alone.
- Distinguish local command discovery from live cloud provisioning.
- Keep destructive cleanup, live-deployment, and credentialed operations out of the default path unless the user explicitly asks for them.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp 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.
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.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
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()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert 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.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match 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).