service-integrations
Operate M-flow as a FastAPI service, web UI, MCP server, Docker Compose stack, authenticated multi-tenant backend, cloud sync/distributed worker, or face-aware playground integration.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- service-integrations
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Operate M-flow as a FastAPI service, web UI, MCP server, Docker Compose stack, authenticated multi-tenant backend, cloud sync/distributed worker, or face-aware playground integration.
- disable-model-invocation
- true
02 资源文件 6
- 参考文档references/api-server-reference.md4.5 KB
- 参考文档references/deployment-and-integrations.md4.0 KB
- 参考文档references/mcp-reference.md3.8 KB
- 参考文档references/troubleshooting.md3.6 KB
- 脚本scripts/mcp_tool_summary.py5.5 KB
- 脚本scripts/service_status_check.py8.1 KB
03 SKILL.md 正文
service-integrations
Use this sub-skill for M-flow service surfaces: API startup, Docker Compose, the web UI, MCP transport and API mode, auth and settings, cloud sync, distributed workers, and the face-aware playground.
Route elsewhere
- In-process Python calls such as
m_flow.add(),memorize(),search(),query(),ingest(), orlearn()→../core-memory-api/ - Storage backend and retrieval tuning details →
../retrieval-graph-search/ - Loaders, chunking, content routing, and pipeline internals →
../ingestion-pipelines/
Use this sub-skill when
- starting or checking
m_flow.api.client:app - launching
mflow -uiorm_flow.api.v1.ui.start_ui() - wiring
m_flow-frontend,m_flow-mcp, Docker Compose, or the playground stack - configuring auth, multi-tenancy, settings, or health checks
- using MCP stdio/SSE/HTTP, remote API mode, or task tracking
- enabling cloud sync or Modal workers
- connecting face-aware playground sessions to
fanjing-face-recognition
Bundled files
- deployment-and-integrations.md
- api-server-reference.md
- mcp-reference.md
- troubleshooting.md
- service_status_check.py
- mcp_tool_summary.py
Guardrails
- Prefer safe status checks over service mutation.
- Treat face-recognition playground setup as distilled guidance only; it downloads models and needs camera access, so do not run automated installers unless the user explicitly asks.
- Do not route core memory operations here; use the core-memory API sub-skill instead.
- Keep all runtime links inside this sub-skill tree or to sibling generated skill files.
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).