AGENT SKILL

object-storage-architect

Architect-level object storage design — bucket and prefix layout, naming and immutability, lifecycle and storage classes, access patterns and presigned URLs, versioning and replication, and the egress and request costs that surprise teams. Use when files, media, backups, or data-lake content live in object storage.

License 许可
MIT
最近更新
2026-11-13
Stars
★ 0
Forks
0

01 前言 Front Matter

name
object-storage-architect
license
MIT
metadata
{ "crusor-path": "cloud-and-devops-architect/data-services-architect/object-storage-architect", "crusor-domain": "cloud-and-devops-architect", "crusor-origin": "original", "crusor-sources": "docs.aws.amazon.com/AmazonS3/latest/userguide; cloud.google.com/storage/docs; learn.microsoft.com/azure/storage/blobs", "crusor-triggers": "s3, object storage, bucket, presigned url, lifecycle policy, storage class, glacier, egress cost, gcs, blob storage", "crusor-verified-on": "2026-10-30" }
description
Architect-level object storage design — bucket and prefix layout, naming and immutability, lifecycle and storage classes, access patterns and presigned URLs, versioning and replication, and the egress and request costs that surprise teams. Use when files, media, backups, or data-lake content live in object storage.
compatibility
S3, GCS, Azure Blob, R2

03 SKILL.md 正文

Object Storage Architect

Object storage is a namespace with a billing model attached. The layout decides what you can delete, replicate, and afford.

When to use

  • User uploads, media, documents, exports.
  • Data lake landing zones, backups, logs archive.
  • Static site assets and build artefacts.

When NOT to use

  • Anything needing partial update or transactions → a database.
  • A filesystem with random writes → block storage or EFS/managed NFS.
  • Serving small hot files repeatedly → put a CDN in front.

Decision 1 — Layout

<bucket-per-domain>
  env=<prod>/<dataset>/dt=2026-10-30/part-*.parquet     analytics
  <tenant>/<entity>/<uuid>/<filename>                    user content
  _meta/manifests/, _quarantine/                          pipeline metadata
  • Prefixes are not directories: listing cost and partition pruning depend on a deliberate key layout.
  • One bucket per purpose and environment; a shared bucket mixes permissions, retention, and billing.
  • Never encode mutable state in the key — keys are permanent.
  • Content-address (hash) for deduplication of blobs; uuid for user-owned assets.

Decision 2 — Access

  • Block public access by default at the account level; public assets are served through a CDN origin, not a public bucket.
  • Workload access by role; presigned URLs for user downloads with a short expiry and a content-disposition.
  • Bucket policies scoped by prefix and principal; deny-by-default elsewhere.
  • Encryption at rest (SSE-KMS for sensitive data) and in transit enforced by policy.
  • Cross-account access by bucket policy + role, never by copying credentials.

Decision 3 — Durability and versioning

  • Versioning on for anything irreplaceable; it also protects against ransomware and accidental deletes.
  • Object lock for compliance retention; understand that it is irreversible until expiry.
  • Cross-region replication for DR; define RPO and verify replication lag.
  • MFA delete for the buckets that hold the backups of everything else.

Decision 4 — Lifecycle and class

Data Class progression
hot user content standard, no transition
warm analytics standard → infrequent (30 d) → archive (180 d)
logs infrequent (7 d) → expire (90 d)
backups archive after 30 d, expire after retention
  • Lifecycle rules are the cheapest cost control in cloud; write them per bucket at creation.
  • Incomplete multipart uploads must expire — they are invisible cost.
  • Request costs matter at high volume (millions of small GETs); batch and cache.

Decision 5 — Pipeline interaction

  • Landing zone is append-only; downstream jobs read a manifest, not a raw listing.
  • Atomic publication: write to a temp prefix, then move/rename via a manifest commit.
  • Schema and partition layout documented once, enforced by the writer.
  • Quality checks on ingest; quarantine prefix for rejects.

Failure modes at scale

  • Public bucket discovered by a scanner; every object world-readable.
  • Lifecycle rules absent; five years of logs in standard storage.
  • Egress from object storage to the internet cheaper to serve from a CDN, ignored.
  • Key layout by uuid only, so listing a tenant's files requires a full scan.
  • Versioning off; a bad deploy overwrites the only copy of user uploads.

Verification receipts

  1. Bucket inventory: purpose, public-access block, encryption, versioning, owner.
  2. Lifecycle policy per bucket with the expected monthly saving.
  3. Access test: presigned URL expires; direct anonymous GET denied.
  4. Replication lag measured for DR buckets.
  5. Cost report split by storage, requests, and egress.

04 同仓库更多 Skills

SriAkhilSJ
SriAkhilSJ/PulseAIRepo

interaction-patterns-architect

Architect-level interaction design in code — choosing established patterns over invented ones, feedback and latency behaviour, destructive-action safeguards, navigation and overlay conventions, and the state model behind every interactive surface. Use when a screen has to behave, not just look.

★ 0
SriAkhilSJ
SriAkhilSJ/PulseAIRepo

design-system-ops-architect

Architect-level operation of a design system in code — token pipelines and versioning, component API and variant policy, the catalogue as the contract, adoption metrics, and the governance that stops a system from forking into three. Use when a component library must survive several teams.

★ 0
SriAkhilSJ
SriAkhilSJ/PulseAIRepo

design-handoff-architect

Architect-level design handoff — what must be specified before implementation starts, tokens over raw values, responsive and content-limit contracts, state matrices, and the review loop that catches drift between design and code. Use when a design is about to become code.

★ 0
SriAkhilSJ
SriAkhilSJ/PulseAIRepo

content-microcopy-architect

Architect-level UI content — where strings live, voice and terminology consistency, error and empty-state copy as part of the interface, translation-ready authoring, and the content model a product needs before it multiplies screens. Use when the words in the interface matter as much as the pixels.

★ 0
SriAkhilSJ
SriAkhilSJ/PulseAIRepo

ui-ux-implementation-architect

Architect-level direction for turning a design into working UI — the handoff contract, component ownership versus screen-level layout, interaction and state conventions, design QA gates, and the review habits that keep an implementation faithful without frozen pixels. Use before any specific UI implementation skill.

★ 0
SriAkhilSJ
SriAkhilSJ/PulseAIRepo

test-strategy-architect

Architect-level test strategy — allocating assertions across unit, integration, contract and end-to-end layers, designing for testability through dependency boundaries, fixture and data strategy, and the flake and coverage policies that keep a suite trustworthy. Use when defining how a codebase is tested.

★ 0