secrets-and-keys-architect
Architect-level secrets management — secret stores and their hierarchy, rotation without downtime, envelope encryption with KMS/HSM, key custody and destruction, environment and CI secret injection, and the leak-detection path that catches what escapes. Use when credentials, tokens, or encryption keys must be handled by systems.
- License 许可
- MIT
- 最近更新
- 2026-11-13
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- secrets-and-keys-architect
- license
- MIT
- metadata
- { "crusor-path": "cloud-and-devops-architect/identity-architect/secrets-and-keys-architect", "crusor-domain": "cloud-and-devops-architect", "crusor-origin": "original", "crusor-sources": "docs.aws.amazon.com/secretsmanager; cloud.google.com/secret-manager/docs; developer.hashicorp.com/vault/docs", "crusor-triggers": "secrets management, vault, kms, key rotation, envelope encryption, secret manager, api keys, key custody, hsm, leaked secret", "crusor-verified-on": "2026-10-30" }
- description
- Architect-level secrets management — secret stores and their hierarchy, rotation without downtime, envelope encryption with KMS/HSM, key custody and destruction, environment and CI secret injection, and the leak-detection path that catches what escapes. Use when credentials, tokens, or encryption keys must be handled by systems.
- compatibility
- AWS Secrets Manager/KMS, GCP Secret Manager, Azure Key Vault, Vault, SOPS
03 SKILL.md 正文
Secrets & Keys Architect
Secrets fail in two ways: they leak, or they cannot be rotated. Every design decision here is about making both cheap.
When to use
- Application credentials, API keys, signing keys, database passwords.
- Encryption at rest with customer-managed keys.
- A leak has happened or is suspected.
When NOT to use
- Human identity and permissions →
iam-architect. - Mobile/web client token storage → mobile or web security books.
Decision 1 — Store
| Store | Fit |
|---|---|
| cloud secret manager | default; IAM-scoped, audited, versioned |
| Vault | multi-cloud, dynamic credentials, PKI |
| KMS/HSM | encryption keys, never raw app secrets |
| SOPS + repo | config encrypted at rest, bootstrapped by a key |
| env vars / files | delivery mechanism only, never the store |
- Secrets never live in git, images, Terraform state in plaintext, or CI logs.
- One secret per credential per environment; sharing a secret across services makes rotation a coordination project.
Decision 2 — Access
- Workload identity reads its own secrets; no human reads production secrets routinely.
- Path/namespace per service, policy scoped to the identity.
- Every read is audited; alert on access from an unexpected identity or region.
- Prefer short-lived dynamic credentials (Vault database engine, IAM auth) over static passwords where the system supports it.
Decision 3 — Rotation
- Rotation is designed in, with two valid credentials during cutover.
- Automated rotation where the provider supports it (database, cloud keys); scheduled and monitored where it does not.
- Consumers re-read the secret or watch for change; a restart requirement is an outage waiting for a schedule.
- Test rotation in staging on every change of the rotation job — a broken rotator is discovered at 3 a.m. otherwise.
Decision 4 — Encryption keys
- Envelope encryption: data key per object/tenant, wrapped by a KMS key.
- Key hierarchy: root/HSM → KMS key → data keys; permissions narrow at each level.
- Separate keys per environment and, where required, per tenant (crypto-shredding becomes deleting one key).
- Key destruction is irreversible: require a waiting period, dual approval, and a backup decision recorded.
Decision 5 — Delivery to workloads
- Injected at runtime via the platform's identity (IRSA, workload identity, managed identity), not baked into images.
- CI/CD gets short-lived credentials by OIDC; no long-lived deploy keys.
- Local development uses a dev-only store; never a copy of production secrets.
- Secret changes do not require an image rebuild.
Decision 6 — Detection
- Scan repos and images for secret patterns; block the merge, not just alert.
- Alert on a secret appearing in logs or error trackers.
- Leak response runbook: revoke → rotate → audit access → notify. Revoke first.
- Assume a leaked secret was used; check the audit trail.
Failure modes at scale
- Database password in a Helm values file, committed and rotated never.
- KMS key deleted to "clean up"; encrypted backups unrecoverable.
- Rotation job fails silently for a year, then the password expires.
- One shared secret across 30 services; rotation requires a coordinated freeze.
- Secret in a CI log line, discovered months later by a third party.
Verification receipts
- Secret inventory: name, owner, consumers, rotation cadence.
- Rotation executed live with zero downtime observed.
- KMS key hierarchy diagram with policies per level.
- Leak scan wired into CI, proven by a deliberate test secret.
- Leak drill: revoke and rotate completed within the target time.
04 同仓库更多 Skills
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.
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.
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.
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.
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.
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.