Expert coverage of Google Cloud Pub/Sub managed messaging service: topics, subscriptions (pull, push, BigQuery, Cloud Storage), ordering keys, exactly-once delivery, dead-letter topics, schemas, snapshots, seek, and monitoring. Use for \"GCP Pub/Sub\", \"Google Pub/Sub\", \"Cloud Pub/Sub\", \"Pub/Sub topic\", \"Pub/Sub subscription\", \"pull subscription\", \"push subscription\", \"BigQuery subscription\", \"Cloud Storage subscription\", \"ordering key\", \"exactly-once Pub/Sub\", \"dead lett...
Pro scans all 4 files and shows the line behind each finding
Scanned 9/24/2026
npx -y skills add chrishuffman5/domain-expert --skill gcp-pubsub --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Gcp Pubsub?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/chrishuffman5-gcp-pubsub)More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.
---
name: gcp-pubsub
description: "Expert coverage of Google Cloud Pub/Sub managed messaging service: topics, subscriptions (pull, push, BigQuery, Cloud Storage), ordering keys, exactly-once delivery, dead-letter topics, schemas, snapshots, seek, and monitoring. Use for \"GCP Pub/Sub\", \"Google Pub/Sub\", \"Cloud Pub/Sub\", \"Pub/Sub topic\", \"Pub/Sub subscription\", \"pull subscription\", \"push subscription\", \"BigQuery subscription\", \"Cloud Storage subscription\", \"ordering key\", \"exactly-once Pub/Sub\", \"dead letter topic\", \"DLT\", \"Pub/Sub schema\", \"Pub/Sub snapshot\", \"seek\", \"gcloud pubsub\", \"message filter Pub/Sub\", \"ack deadline\", \"Pub/Sub flow control\", \"streaming pull\". Do NOT use for general GCP architecture or non-messaging resource provisioning -- that's the `cloud-platforms` plugin."
license: MIT
---
# GCP Pub/Sub
This skill covers Google Cloud Pub/Sub, a fully managed serverless messaging service, including:
- Topics (message ordering, schema validation, retention, CMEK encryption)
- Subscription types: Pull (streaming pull), Push (HTTPS webhooks), BigQuery export, Cloud Storage export
- Message ordering (ordering keys, per-key throughput, hot key mitigation)
- Exactly-once delivery (pull subscriptions only, regional requirements)
- Dead-letter topics (DLT with max delivery attempts, IAM requirements)
- Message filtering (attribute-based immutable filters)
- Schema validation (Avro, Protocol Buffers, evolution, revisions)
- Snapshots and seek (point-in-time replay, backlog purge)
- Flow control (subscriber and publisher side)
- Monitoring (Cloud Monitoring metrics, delivery latency health score)
- IAM, VPC Service Controls, CMEK
## How to Approach Tasks
1. **Classify** the request:
- **Architecture / design** -- Load `references/architecture.md` for topics, subscription types, ordering, exactly-once, DLT, schemas
- **Best practices** -- Load `references/best-practices.md` for subscription selection, ordering key design, ack deadline tuning, flow control, cost optimization
- **Troubleshooting** -- Load `references/diagnostics.md` for unacked buildup, delivery latency, DLT analysis, push endpoint errors
2. **Recommend** -- Provide actionable guidance with `gcloud pubsub` commands, Terraform examples, and SDK patterns.
## Core Architecture
### Per-Message Leasing (Not Partitioned)
Each message is individually leased to a subscriber. No partition management. Any subscriber can receive any message. Serverless auto-scaling.
### Topics
Named resource for publishing. Optional: message retention (10 min to 31 days), ordering (immutable at creation), schema validation, CMEK.
### Subscription Types
- **Pull:** Client requests messages. Best for high throughput, exactly-once. Streaming pull for lowest latency.
- **Push:** Pub/Sub sends HTTP POST to HTTPS endpoint. Best for Cloud Run, Cloud Functions.
- **BigQuery:** Direct export via Storage Write API. Schema mapping.
- **Cloud Storage:** Batch messages into files (text or Avro).
### Dead-Letter Topics
Messages forwarded after 5-100 delivery attempts. Requires IAM permissions for Pub/Sub service account.
## Anti-Patterns
| Anti-Pattern | Why It Fails | Instead |
|---|---|---|
| Ordering with low-cardinality keys | Hot keys block progress; 1 MBps per key limit | Use high-cardinality keys (per entity) |
| Acking before processing complete | Recovery requires seek; messages lost | Ack only after successful processing |
| Push to unreliable endpoint | 5xx triggers exponential backoff; delivery stalls | Use pull for unreliable consumers; fix endpoint |
| Large message retention without monitoring | Storage costs accumulate silently | Set appropriate retention; monitor orphaned subs |
| Ignoring ack deadline expiration | Messages redelivered; duplicate processing | Match ack deadline to P99 processing time |
## Reference Files
- `references/architecture.md` -- Topics, subscription types (pull/push/BigQuery/Cloud Storage), ordering keys, exactly-once, DLT, schemas, snapshots/seek, flow control
- `references/best-practices.md` -- Subscription selection, ordering key design, ack deadline tuning, flow control, retry policies, schema evolution, cost optimization
- `references/diagnostics.md` -- Unacked message buildup, delivery latency health score, DLT analysis, push endpoint errors, Cloud Monitoring metrics, alerting setup
## Cross-References
- The `overview` skill -- cross-broker comparisons and event-driven architecture design
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!