viktor/infra
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
infra/.claude/CLAUDE.md

27 lines
2.1 KiB
Markdown
Raw Normal View History

[ci skip] add AGENTS.md for model-agnostic knowledge, slim CLAUDE.md to Claude-specific layer AGENTS.md (63 lines): shared infra knowledge for any AI tool (Codex, Claude, Cursor). Covers: critical rules, architecture, storage, tiers, common ops. CLAUDE.md (23 lines): Claude-specific addons — skills, agents, user preferences. References AGENTS.md for shared knowledge. Removed generic agents (devops-engineer, fullstack-developer).
2026-03-06 23:50:26 +00:00
# Claude Code — Project Configuration
add prometheus alerts for deployment/statefulset/daemonset replica mismatches [ci skip] - Add DeploymentReplicasMismatch alert - Add StatefulSetReplicasMismatch alert - Add DaemonSetMissingPods alert - Add .claude/ directory with remote executor and knowledge base
2026-01-18 11:04:51 +00:00
[ci skip] add AGENTS.md for model-agnostic knowledge, slim CLAUDE.md to Claude-specific layer AGENTS.md (63 lines): shared infra knowledge for any AI tool (Codex, Claude, Cursor). Covers: critical rules, architecture, storage, tiers, common ops. CLAUDE.md (23 lines): Claude-specific addons — skills, agents, user preferences. References AGENTS.md for shared knowledge. Removed generic agents (devops-engineer, fullstack-developer).
2026-03-06 23:50:26 +00:00
> **Shared knowledge**: Read `AGENTS.md` at repo root for architecture, patterns, rules, and operations. This file adds Claude-specific features on top.
[ci skip] refactor claude files: compact CLAUDE.md, clean memory, remove generic agents CLAUDE.md: 260→72 lines. Moved detailed patterns (NFS, iSCSI, Kyverno tables, anti-AI, node rebuild) to .claude/reference/patterns.md. Kept: critical rules, quick patterns, key commands, tier overview, prefs. Memory: CLAUDE.md is now single source of truth. Auto-memory reduced to scratch pad (67→25 lines, 5→1 files). MetaClaw DB cleaned from 40→16 entries (removed all infra-specific duplicates, kept cross-project prefs). Agents: removed generic devops-engineer (885L) and fullstack-developer (234L). Kept custom cluster-health-checker (48L).
2026-03-06 23:27:46 +00:00
[ci skip] add AGENTS.md for model-agnostic knowledge, slim CLAUDE.md to Claude-specific layer AGENTS.md (63 lines): shared infra knowledge for any AI tool (Codex, Claude, Cursor). Covers: critical rules, architecture, storage, tiers, common ops. CLAUDE.md (23 lines): Claude-specific addons — skills, agents, user preferences. References AGENTS.md for shared knowledge. Removed generic agents (devops-engineer, fullstack-developer).
2026-03-06 23:50:26 +00:00
## Claude-Specific Resources
- **Skills**: `.claude/skills/` (7 active). Archived runbooks: `.claude/skills/archived/`
- **Agents**: `.claude/agents/cluster-health-checker` (haiku, autonomous health checks)
- **Reference**: `.claude/reference/` — patterns.md, service-catalog.md, proxmox-inventory.md, github-api.md, authentik-state.md
[ci skip] refactor claude files: compact CLAUDE.md, clean memory, remove generic agents CLAUDE.md: 260→72 lines. Moved detailed patterns (NFS, iSCSI, Kyverno tables, anti-AI, node rebuild) to .claude/reference/patterns.md. Kept: critical rules, quick patterns, key commands, tier overview, prefs. Memory: CLAUDE.md is now single source of truth. Auto-memory reduced to scratch pad (67→25 lines, 5→1 files). MetaClaw DB cleaned from 40→16 entries (removed all infra-specific duplicates, kept cross-project prefs). Agents: removed generic devops-engineer (885L) and fullstack-developer (234L). Kept custom cluster-health-checker (48L).
2026-03-06 23:27:46 +00:00
- **GitHub API**: `curl` with tokens from tfvars (`gh` CLI blocked by sandbox)
[ci skip] add AGENTS.md for model-agnostic knowledge, slim CLAUDE.md to Claude-specific layer AGENTS.md (63 lines): shared infra knowledge for any AI tool (Codex, Claude, Cursor). Covers: critical rules, architecture, storage, tiers, common ops. CLAUDE.md (23 lines): Claude-specific addons — skills, agents, user preferences. References AGENTS.md for shared knowledge. Removed generic agents (devops-engineer, fullstack-developer).
2026-03-06 23:50:26 +00:00
## Instructions
[ci skip] update AGENTS.md + CLAUDE.md with SOPS workflow, add k8s-portal CI pipeline AGENTS.md: added SOPS secrets management section, scripts/tg usage, contributor onboarding steps, pull-through cache bypass notes. CLAUDE.md: added SOPS workflow note, linux/amd64 build reminder, versioned tag guidance for pull-through cache. CI: new .woodpecker/k8s-portal.yml pipeline — auto-builds and deploys the k8s portal when files under stacks/platform/modules/k8s-portal/files/ change on master push. Uses buildx for linux/amd64.
2026-03-07 15:37:19 +00:00
- **"remember X"**: Update this file + `AGENTS.md` (if shared knowledge), commit with `[ci skip]`
- **Apply with SOPS**: Use `scripts/tg` wrapper instead of raw `terragrunt` — auto-decrypts secrets
[ci skip] add AGENTS.md for model-agnostic knowledge, slim CLAUDE.md to Claude-specific layer AGENTS.md (63 lines): shared infra knowledge for any AI tool (Codex, Claude, Cursor). Covers: critical rules, architecture, storage, tiers, common ops. CLAUDE.md (23 lines): Claude-specific addons — skills, agents, user preferences. References AGENTS.md for shared knowledge. Removed generic agents (devops-engineer, fullstack-developer).
2026-03-06 23:50:26 +00:00
- **New services need CI/CD** (Woodpecker) and **monitoring** (Prometheus/Uptime Kuma)
- **New service**: Use `setup-project` skill for full workflow
[ci skip] refactor claude files: compact CLAUDE.md, clean memory, remove generic agents CLAUDE.md: 260→72 lines. Moved detailed patterns (NFS, iSCSI, Kyverno tables, anti-AI, node rebuild) to .claude/reference/patterns.md. Kept: critical rules, quick patterns, key commands, tier overview, prefs. Memory: CLAUDE.md is now single source of truth. Auto-memory reduced to scratch pad (67→25 lines, 5→1 files). MetaClaw DB cleaned from 40→16 entries (removed all infra-specific duplicates, kept cross-project prefs). Agents: removed generic devops-engineer (885L) and fullstack-developer (234L). Kept custom cluster-health-checker (48L).
2026-03-06 23:27:46 +00:00
- **Ingress**: `ingress_factory` module. Auth: `protected = true`. Anti-AI: on by default.
[ci skip] update AGENTS.md + CLAUDE.md with SOPS workflow, add k8s-portal CI pipeline AGENTS.md: added SOPS secrets management section, scripts/tg usage, contributor onboarding steps, pull-through cache bypass notes. CLAUDE.md: added SOPS workflow note, linux/amd64 build reminder, versioned tag guidance for pull-through cache. CI: new .woodpecker/k8s-portal.yml pipeline — auto-builds and deploys the k8s portal when files under stacks/platform/modules/k8s-portal/files/ change on master push. Uses buildx for linux/amd64.
2026-03-07 15:37:19 +00:00
- **Docker images**: Always build for `linux/amd64` (`docker buildx build --platform linux/amd64`). Pull-through cache serves stale :latest — use versioned tags.
[ci skip] remember: update kubelet thresholds when changing node memory
2026-03-08 10:34:17 +00:00
- **Node memory changes**: When changing VM memory on any k8s node, update kubelet `systemReserved`, `kubeReserved`, and eviction thresholds accordingly. Config: `/var/lib/kubelet/config.yaml`. Template: `stacks/infra/main.tf`. Current values: systemReserved=512Mi, kubeReserved=512Mi, evictionHard=500Mi, evictionSoft=1Gi.
[ci skip] add sealed secrets convention: fileset + kubernetes_manifest pattern - Document sealed secrets workflow in AGENTS.md and CLAUDE.md - Add kubernetes_manifest + fileset(sealed-*.yaml) block to plotting-book as reference - Users: kubeseal encrypt → commit sealed-*.yaml → CI applies via Terraform - E2E tested: seal/commit/plan/apply/decrypt cycle verified
2026-03-08 20:03:50 +00:00
- **Sealed Secrets**: User-managed secrets go in `sealed-*.yaml` files in the stack directory. Stacks pick them up via `kubernetes_manifest` + `fileset(path.module, "sealed-*.yaml")`. See AGENTS.md for full workflow.
[ci skip] refactor claude files: compact CLAUDE.md, clean memory, remove generic agents CLAUDE.md: 260→72 lines. Moved detailed patterns (NFS, iSCSI, Kyverno tables, anti-AI, node rebuild) to .claude/reference/patterns.md. Kept: critical rules, quick patterns, key commands, tier overview, prefs. Memory: CLAUDE.md is now single source of truth. Auto-memory reduced to scratch pad (67→25 lines, 5→1 files). MetaClaw DB cleaned from 40→16 entries (removed all infra-specific duplicates, kept cross-project prefs). Agents: removed generic devops-engineer (885L) and fullstack-developer (234L). Kept custom cluster-health-checker (48L).
2026-03-06 23:27:46 +00:00
add claude [ci skip]
2026-02-06 20:10:02 +00:00
## User Preferences
[ci skip] refactor claude files: compact CLAUDE.md, clean memory, remove generic agents CLAUDE.md: 260→72 lines. Moved detailed patterns (NFS, iSCSI, Kyverno tables, anti-AI, node rebuild) to .claude/reference/patterns.md. Kept: critical rules, quick patterns, key commands, tier overview, prefs. Memory: CLAUDE.md is now single source of truth. Auto-memory reduced to scratch pad (67→25 lines, 5→1 files). MetaClaw DB cleaned from 40→16 entries (removed all infra-specific duplicates, kept cross-project prefs). Agents: removed generic devops-engineer (885L) and fullstack-developer (234L). Kept custom cluster-health-checker (48L).
2026-03-06 23:27:46 +00:00
- **Calendar**: Nextcloud at `nextcloud.viktorbarzin.me`
- **Home Assistant**: ha-london (default), ha-sofia. "ha"/"HA" = ha-london
[ci skip] Refactor knowledge: CLAUDE.md 881→190 lines, extract reference data CLAUDE.md changes: - Extract service catalog + Cloudflare domains → .claude/reference/service-catalog.md - Extract Proxmox VMs, hardware, network → .claude/reference/proxmox-inventory.md - Extract GitHub/Drone API patterns → .claude/reference/github-drone-api.md - Extract Authentik state snapshot → .claude/reference/authentik-state.md - Remove Init Container pattern (duplicates setup-project skill) - Remove Poison Fountain service notes (duplicates Anti-AI section) - Consolidate Authentik section (link to skills + reference) - Remove resource limit tables (kept tier definitions inline) Skill merges (37→32): - helm-release-force-rerender + helm-stuck-release-recovery → helm-release-troubleshooting - containerd-multi-registry-pull-through-cache + k8s-docker-registry-cache-bypass → k8s-container-image-caching - (traefik merges in previous commits)
2026-02-22 22:11:31 +00:00
- **Frontend**: Svelte for all new web apps
[ci skip] refactor claude files: compact CLAUDE.md, clean memory, remove generic agents CLAUDE.md: 260→72 lines. Moved detailed patterns (NFS, iSCSI, Kyverno tables, anti-AI, node rebuild) to .claude/reference/patterns.md. Kept: critical rules, quick patterns, key commands, tier overview, prefs. Memory: CLAUDE.md is now single source of truth. Auto-memory reduced to scratch pad (67→25 lines, 5→1 files). MetaClaw DB cleaned from 40→16 entries (removed all infra-specific duplicates, kept cross-project prefs). Agents: removed generic devops-engineer (885L) and fullstack-developer (234L). Kept custom cluster-health-checker (48L).
2026-03-06 23:27:46 +00:00
- **Tools**: Docker containers only — never `brew install` locally
- **Pod monitoring**: Never use `sleep` — spawn background subagent with `kubectl get pods -w`
Reference in a new issue Copy permalink