ui-and-storage
Routes UltraRAG UI, chat, knowledge-base, memory, and storage workflows backed by the Flask application.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- ui-and-storage
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Routes UltraRAG UI, chat, knowledge-base, memory, and storage workflows backed by the Flask application.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/storage-auth.md2.2 KB
- 参考文档references/troubleshooting.md2.3 KB
- 参考文档references/ui-backend.md3.9 KB
- 脚本scripts/inspect_ui_backend.py2.9 KB
03 SKILL.md 正文
UI and Storage
Use this sub-skill when the task is about the UltraRAG web UI, local storage, chat sessions, knowledge-base operations, or the case-study viewer.
Typical triggers
ultrarag show uiorultrarag show caseui/backend,create_app, pipeline manager, auth, chat store, KB visibility- chat sessions, background chats, export, or memory sync
- uploading, indexing, deleting, or inspecting KB files
ULTRARAG_UI_STORAGE_ROOT,ULTRARAG_FRONTEND_DIR, or session timeout env vars- Frontend build or asset serving questions
What this sub-skill covers
- The Flask backend created by
ui/backend/app.py. - Storage layout and the SQLite-backed auth/chat/visibility stores.
- Knowledge-base CRUD, pipeline CRUD, background chat tasks, and memory sync.
- Frontend asset resolution and the case-study viewer data format.
What stays elsewhere
- Pipeline YAML and orchestration belong in
sub-skills/pipelines/. - Server tool/prompt contracts and backend selection belong in
sub-skills/servers/.
Start here
- Read
references/ui-backend.mdfor the route catalog and main Flask entry points. - Read
references/storage-auth.mdfor the storage tree, SQLite stores, and environment variables. - Read
references/troubleshooting.mdwhen the UI, case-study viewer, or KB storage fails. - Run
scripts/inspect_ui_backend.pyfor a quick route and storage summary.
Common user questions this sub-skill should answer
- How do I start the UI or case-study viewer?
- Which directories and environment variables control UI storage?
- How do chat sessions, KB files, and visibility mappings get stored?
- Why is
show uiorshow casefailing on this machine? - How do I inspect the Flask route set without reopening the source tree?
Practical workflow
- Confirm whether the user wants the UI server, the case-study viewer, or the storage back end.
- Check the storage root and the frontend asset directory.
- Use the route and storage references to pick the right fix.
- If the issue is actually with pipeline semantics or a server backend, switch to the corresponding sub-skill.
Helpful commands
ultrarag show ui --host 127.0.0.1 --port 5050
ultrarag show case --config_path output/memory_*.json
If the user is asking about chat logic inside the pipeline engine itself, go back to the pipelines sub-skill instead of editing UI storage rules here.
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).