automation-schedules-sensors
Define, debug, and test Dagster schedules, sensors, asset sensors, run status sensors, declarative automation conditions, freshness automation, and backfill policies.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- automation-schedules-sensors
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Define, debug, and test Dagster schedules, sensors, asset sensors, run status sensors, declarative automation conditions, freshness automation, and backfill policies.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/api-reference.md8.1 KB
- 参考文档references/troubleshooting.md8.1 KB
- 参考文档references/workflows.md7.0 KB
- 脚本scripts/preview_sensor_schedule.py5.1 KB
03 SKILL.md 正文
Automation: schedules, sensors, and declarative conditions
Use this sub-skill when the task is about time-based or event-based automation in Dagster code:
ScheduleDefinition,@schedule, cron schedules, time zones, schedule run config/tags, and schedule unit tests.@sensor,RunRequest,SkipReason,SensorResult, cursor handling, run keys, resource-backed sensors, and sensor unit tests.@asset_sensor, run status sensors, failure sensors, and tests using built Dagster context objects.AutomationCondition,AutomationConditionSensorDefinition, freshness-driven automation, andBackfillPolicyfor partitioned assets.
Route elsewhere when the task is primarily about core asset modeling, deployment/daemon operations, or starting/stopping local CLI processes.
Start here
- For implementation patterns and API choices, read references/api-reference.md.
- For end-to-end workflows and test commands, read references/workflows.md.
- For common failures and fixes, read references/troubleshooting.md.
- To safely inspect what CLI preview commands would be used, run
python skills/dagster/sub-skills/automation-schedules-sensors/scripts/preview_sensor_schedule.py --help.
Quick routing
- Need a cron trigger for a job or asset job: use
@scheduleorScheduleDefinition; test withbuild_schedule_context()and.evaluate_tick(). - Need polling logic or external-event logic: use
@sensor, stableRunRequest(run_key=...), andcontext.update_cursor(...)orSensorResult(cursor=...). - Need to react to one asset materialization: use
@asset_sensor; prefer declarative automation for asset dependency propagation when custom per-event run config is not required. - Need to react to job success/failure: use
@run_status_sensoror@run_failure_sensor; avoid triggering an infinite loop when the requested job can produce the same monitored status. - Need asset-centric freshness, dependency, or backfill behavior: prefer
AutomationCondition.on_cron(),AutomationCondition.eager(),AutomationCondition.on_missing(), freshness policies/checks, andBackfillPolicy.
Definition checklist
- Put schedules and sensors in the repository or
Definitions(schedules=[...], sensors=[...])that loads with the target jobs/assets. - Use only one target style per schedule/sensor:
job,job_name,jobs,asset_selection, ortargetas accepted by that definition. - Give every multi-request schedule a non-empty
run_keyon eachRunRequest; use stable sensor run keys only when duplicate suppression should survive cursor resets. - Keep schedule/sensor evaluation pure enough for unit tests: move network/credentialed calls behind resources that can be stubbed in
build_*_context. - For declarative automation, ensure the automation condition sensor is present and enabled in the deployment; route daemon lifecycle and deployment checks to the deployment operations 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).