api-client-settings
Use Prefect Python clients, settings, profiles, schemas, and the prefect-client package split safely.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- api-client-settings
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Use Prefect Python clients, settings, profiles, schemas, and the prefect-client package split safely.
- disable-model-invocation
- true
02 资源文件 5
- 参考文档references/api-reference.md5.2 KB
- 参考文档references/prefect-client-package.md4.0 KB
- 参考文档references/settings-and-profiles.md5.2 KB
- 参考文档references/troubleshooting.md4.9 KB
- 脚本scripts/inspect_prefect_settings.py6.1 KB
03 SKILL.md 正文
Prefect API Client and Settings
Use this sub-skill when a task needs direct Python access to Prefect orchestration or Cloud APIs, settings/profile diagnosis, schema-model construction, or the lightweight prefect-client package boundary.
Start Here
- Choose the client style in references/api-reference.md:
async with get_client()for async code,with get_client(sync_client=True)for synchronous code. - Confirm configuration in references/settings-and-profiles.md:
PREFECT_API_URL,PREFECT_API_KEY, active profile, and source precedence. - If the environment uses
prefect-clientinstead of the fullprefectpackage, read references/prefect-client-package.md before importing CLI/server modules. - For failures, use references/troubleshooting.md and the bundled
scripts/inspect_prefect_settings.pyhelper.
Use For
get_client,PrefectClient,SyncPrefectClient, and direct REST API method calls.- Pydantic schema objects under
prefect.client.schemas.*for filters, sorts, actions, objects, and responses. get_current_settings,temporary_settings, settings accessors, environment variables,.env,prefect.toml,pyproject.toml, and profiles TOML.prefect-clientpackage behavior and custom generated deployment SDK usage patterns.
Route Elsewhere
- Server implementation internals, generated schema maintenance, and repo test selection:
../repo-development/SKILL.md. - CLI command operations, server startup, Cloud login commands, and profile/config command usage:
../cli-server-operations/SKILL.md. - Flow/task decorators, execution semantics, states, futures, and task runners:
../flow-task-authoring/SKILL.md. - Deployment creation, work pools, workers, and deployment YAML:
../deployments-workers/SKILL.md.
Safe Validation
- Inspect effective settings without network calls:
python scripts/inspect_prefect_settings.py --schema-summary
- Validate a profiles TOML file before relying on it:
python scripts/inspect_prefect_settings.py --validate-profiles --profiles-path ~/.prefect/profiles.toml
- For client code, prefer tiny
hello()or read-only list/filter calls against a known API URL; do not start long-lived services from this sub-skill.
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).