deployment-runtime
Start Gerev from source or Docker, build the UI, and troubleshoot storage, boot, and deployment layout issues.
- License 许可
- MIT
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- deployment-runtime
- license
- MIT
- metadata
- { "disco-role": "operating" }
- description
- Start Gerev from source or Docker, build the UI, and troubleshoot storage, boot, and deployment layout issues.
- disable-model-invocation
- true
02 资源文件 3
- 参考文档references/deployment-runtime.md2.3 KB
- 参考文档references/troubleshooting.md2.0 KB
- 脚本scripts/inspect_runtime_paths.py4.5 KB
03 SKILL.md 正文
Gerev Deployment and Runtime
Use this sub-skill when the task is about running Gerev from source, starting the Docker/compose stack, building the UI bundle, or troubleshooting startup/storage/port behavior.
Best-fit tasks
- Start Gerev locally from the backend source tree.
- Explain the Dockerfile, compose file, runtime ports, and storage mount.
- Build or serve the React UI bundle that the backend serves in production.
- Diagnose startup failures that happen before search or connector workflows can run.
- Understand where the backend stores SQLite, queue, Faiss, BM25, and telemetry files.
Start here
-
Read
deployment-runtime.mdfor the boot sequence, Docker/compose layout, storage paths, and UI build expectations. -
Read
troubleshooting.mdfor startup, storage, model-cache, and UI-serving failures. -
Read
frontend-api.mdwhen you need the backend route contract that the UI expects during boot and polling. -
Run the bundled
inspect_runtime_paths.pyread-only helper for a concise layout summary:python scripts/inspect_runtime_paths.py --app-dir <checkout>/app --jsonAdd
--strictwhen you want the helper to exit nonzero if required runtime files are missing.
High-signal routing checks
run.shruns migrations and starts Uvicorn on port80.Dockerfileinstalls the backend dependencies, pre-caches the model stack, and copies the built UI into/ui.docker-compose.yamlmounts~/.gerev/storage:/opt/storageand shows the optional NVIDIA GPU reservation.app/paths.pyselects/opt/storagein Docker-like mode and a home-directory storage tree otherwise.- The current source tree has a known startup blocker in the search/index import chain:
split_PDF_into_paragraphsis imported but missing.
Boundaries
Included here: source startup, Docker/compose, UI build, storage layout, ports, and the boot-time service behavior that makes the app usable.
Do not use this sub-skill for connector credential setup or search ranking internals unless those details are needed only to explain startup/runtime behavior.
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).