Installs into .claude/skills of the current project.
Are you the author of Playbook Data Engineering Streaming Hardening?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/aniruddhaadak80-playbook-data-engineering-streaming-hardening)
---
name: playbook-data-engineering-streaming-hardening
description: "Correct event-time semantics with bounded state and monitored watermarks."
---
# Playbook: Harden a streaming platform
> Correct event-time semantics with bounded state and monitored watermarks.
**Track:** 🗺️ Data Engineering · **Domain:** Journey Playbooks · **Level:** journey · **~75 min**
**Who this is for:** Data Engineers, Analytics Engineers, Platform Data Teams
## When to Use This Skill
Correct event-time semantics with bounded state and monitored watermarks.
Use it whenever a matching task appears in conversation — the agent loads these instructions on demand.
## Journey Steps
1. Step 1 — Choose windowing and watermarks correctly: start with "Timestamp on event time extracted from payload, not arrival"
2. Step 2 — Make batch pipelines rerunnable by design: start with "Key every write by logical date/partition; overwrite partitions, never append blind"
3. How it fits together: Watermark lag tells the truth; dashboards lag it.
### Referenced Skills
- `streaming-modeling-stream-processing-windows`
- `pipelines-batch-pipeline-idempotency`
## Commands
**Install all referenced skills**
```bash
npx skills add aniruddhaadak80/skills --skill streaming-modeling-stream-processing-windows && npx skills add aniruddhaadak80/skills --skill pipelines-batch-pipeline-idempotency
```
---
Part of [aniruddhaadak80/skills](https://github.com/aniruddhaadak80/skills) · Browse all at https://skills.sh/aniruddhaadak80/skills
