generated from coulomb/repo-seed
docs: introduce CONFIG.md config point registry (CP-NK-001)
Establishes the local config point registry pattern. A config point is an explicit non-default value that cannot be derived from topology, naming convention, component defaults, or automation. Minimizing the list is a design goal — absence of an entry means "accepting the upstream default". CP-NK-001: ACME contact email (bernd.worsch+netkingdom@gmail.com) Location: sso-mfa/k8s/cert-manager/issuers.yaml:38 Why non-default: ACME requires a real monitored inbox; no system default qualifies. Automation via Local Identity GECOS email is deferred. State-hub extension point EP 8e1cda6a registered in custodian domain: "Config point registry: centralized view of explicit non-default configuration values" — proposes ConfigPoint entity type + MCP tools + cross-domain minimization metric. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
47
CONFIG.md
Normal file
47
CONFIG.md
Normal file
@@ -0,0 +1,47 @@
|
||||
# Config Point Registry
|
||||
|
||||
## Philosophy
|
||||
|
||||
net-kingdom is opinionated: defaults, conventions, and automation are preferred at every
|
||||
level. A config point in this file is a **conscious exception** — a value that cannot be
|
||||
derived from the system's topology, naming conventions, component defaults, or available
|
||||
automation.
|
||||
|
||||
**Minimizing this list is a design goal.** Before adding a config point, ask:
|
||||
- Can the value be derived from a naming convention or topology fact?
|
||||
- Can it be auto-generated (e.g. from the Linux user identity, like Local Identity does)?
|
||||
- Is the default provided by the upstream component safe to accept?
|
||||
|
||||
If yes to any of the above, don't add it here.
|
||||
|
||||
---
|
||||
|
||||
## Summary
|
||||
|
||||
| ID | Name | Value | Location(s) |
|
||||
|----|------|-------|-------------|
|
||||
| CP-NK-001 | ACME contact email | `bernd.worsch+netkingdom@gmail.com` | `sso-mfa/k8s/cert-manager/issuers.yaml:38` |
|
||||
|
||||
---
|
||||
|
||||
## CP-NK-001 — ACME contact email
|
||||
|
||||
**Value:** `bernd.worsch+netkingdom@gmail.com`
|
||||
**Set:** 2026-03-02
|
||||
**Set by:** worsch
|
||||
|
||||
**Location(s):**
|
||||
- `sso-mfa/k8s/cert-manager/issuers.yaml:38` — `spec.acme.email` on the
|
||||
`letsencrypt-prod` ClusterIssuer
|
||||
|
||||
**Why non-default:** ACME (Let's Encrypt) requires a contact address for certificate
|
||||
lifecycle notifications — expiry warnings, rate-limit alerts, policy announcements.
|
||||
There is no system-level default that qualifies: this must be a real, monitored inbox.
|
||||
|
||||
**Why not automated:** The Linux user GECOS email (via Local Identity) would be a
|
||||
natural source. However, that introduces a runtime dependency between cluster
|
||||
provisioning and the local-identity tool. Deferred; revisit when Local Identity
|
||||
gains a structured "operator contact" concept.
|
||||
|
||||
**Scope:** All TLS certificates issued by the `letsencrypt-prod` ClusterIssuer across
|
||||
the entire cluster.
|
||||
Reference in New Issue
Block a user