airflow-dag-patterns

Build production Apache Airflow DAGs with best practices for operators, sensors, testing, and deployment. Use when creating data pipelines, orchestrating workflows, or scheduling batch jobs.

Install with Tessl CLI

npx tessl i github:wshobson/agents --skill airflow-dag-patterns

What are skills?

Review — 82%

Does it follow best practices?

If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:

npx tessl skill review --optimize ./path/to/skill

Learn more

Evaluation — 90%

Agent success when using this skill

Validation — 10 / 11 Passed

Validation for skill structure

SKILL.md

Review

Evals

Evaluation results

73%

-17%

Daily Sales Report Pipeline

Production DAG default_args and retry configuration

Criteria

Without context

With context

retries=3

100%

retry_delay 5 min

100%

Exponential backoff

100%

max_retry_delay 1h

100%

email_on_failure True

email_on_retry False

100%

catchup=False

100%

max_active_runs=1

100%

No depends_on_past

100%

Task failure callback

100%

DAG failure callback

100%

No hardcoded date

80%

70%

schedule= syntax

100%

Without context: $0.2329 · 54s · 10 turns · 10 in / 3,569 out tokens

With context: $0.5180 · 1m 37s · 20 turns · 268 in / 5,208 out tokens

100%

Inventory Reconciliation Pipeline with Conditional Processing

Sensor patterns, branching, and trigger rules

Criteria

Without context

With context

TaskFlow @dag decorator

100%

TaskFlow @task decorator

100%

@task.sensor used

100%

Sensor mode=reschedule

100%

Sensor timeout set

100%

Sensor poke_interval set

100%

BranchPythonOperator used

87%

100%

Join trigger rule

100%

Cleanup trigger rule

100%

PokeReturnValue returned

100%

Without context: $0.6989 · 3m 2s · 23 turns · 23 in / 10,764 out tokens

With context: $1.0355 · 3m 48s · 34 turns · 282 in / 12,115 out tokens

98%

Multi-Tenant Report Pipeline with Tests

DAG testing, project structure, and dynamic generation

Criteria

Without context

With context

Factory function

100%

globals() registration

100%

Config-driven clients

100%

Logic in common module

100%

DAG file stays thin

100%

DagBag fixture

100%

include_examples=False

100%

Import error test

100%

Cycle detection test

71%

85%

Structural assertion

100%

Without context: $0.9798 · 3m 30s · 37 turns · 661 in / 10,091 out tokens

With context: $1.5163 · 4m 49s · 51 turns · 48 in / 16,881 out tokens

Evaluated: 3 days ago
Agent: Claude Code

Table of Contents

Daily Sales Report Pipeline Inventory Reconciliation Pipeline with Conditional Processing Multi-Tenant Report Pipeline with Tests

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.