Task-by-task: author ADR-017, expand ADR-008 Level 4, create the VERIFY.md
template + /verify-service skill, and reconcile the checklist/CLAUDE.md/
gitignore/STATUS/TODO. Buildable-now artifacts; live run stays deferred.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Resolves ADR-015 deferred item #2 + TODO 2.2/2.3: a Claude-driven exploratory
browser harness (/verify-service) that exercises staging service UIs through
real SSO, backed by a per-service VERIFY.md, with test users in staging
Authentik and a manual-test handoff. Basis for ADR-017.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Resolves ADR-015 deferred item #1: the mesh VPN is NetBird, self-hosted on
askari, replacing ADR-007's VLAN-99 OPNsense WireGuard. Agent-per-host
enrollment via base, embedded local-user IdP, coordinator off-site for
outage survival. Basis for ADR-016.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Task-by-task docs plan: author ADR-015 and reconcile ADR-001/005/008/009/012,
the new-host and rotate-secrets runbooks, accepted-risks, STATUS, and CLAUDE.md.
Documentation-only; the physical box stays "designed, not built".
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Records the decision to replace the cluster-resident control VM with a
dedicated always-on physical mini-PC (ubongo) outside the Proxmox
cluster, collapsing control plane, AI-worker host, dev home, and local
test runner into one box. Basis for ADR-015.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reviewed the Claude Code config against boma's capabilities and committed a
reproducible, leaner toolchain:
- .claude/settings.json now declares extraKnownMarketplaces + enabledPlugins so a
fresh clone prompts to install the active set: superpowers, context7, terraform
(we use TF, ADR-006), claude-md-management (doc/ADR-heavy). Drops code-simplifier.
- Adds a conservative, read-only/verify permissions allowlist (git status/diff/log,
make lint/test/check, pytest, rbw unlocked, ls/cat/rg/find) — mutations and
outward/destructive commands stay gated, consistent with ADR-002.
- docs/runbooks/claude-code-setup.md: per-machine bootstrap, the deferred
enable-when plugins (security-guidance/semgrep, playwright, hookify, skill-creator),
rbw/venv prerequisites, and a note to keep the dangerous-mode prompt on.
Closes TODO 10.7. Plugin install remains a per-machine /plugin action (no native
auto-install).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A living, high-level map of boma's intended capabilities, organised by domain
(10 domains) with tier tags (platform/app/support) and commitment tags
(core/planned/candidate/maybe-later), mirroring STATUS.md's honesty.
Built capabilities-first on boma's own terms per ADR-013 — V4 was consulted only
as a completeness check (last), not as a source of scope. The check confirmed
strong alignment with V4's actual services, surfaced re-justified gaps (ebook/
audiobook library, Collabora, Samba/NFS, service portal, self-hosted email as
maybe-later), and confirmed deliberate exclusions (V4's desktop/workstation roles;
the dropped Knowledge domain) — boma is server-only by design.
Frames downstream decisions (service picks, Netbird-vs-WireGuard reconcile with
ADR-007, node placement, plugin relevance) without resolving them. Linked from
CLAUDE.md.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Policy for how agents source authoritative knowledge and translate it to boma:
- Facts (version-specific syntax/options) → consult version-matched official docs,
cite, and stamp `verified: subject · tool ver · source · date`. Training memory
is the draft, never the authority.
- Judgments (best practices) → evidence to translate through boma's principles
(extends ADR-013); never adopted because the docs/a blog say so.
- Consultation is risk-based (security / unfamiliar-or-fast-moving / can't quote a
flag with confidence), with a "from memory, unverified" transparency backstop.
- Sources: context7 → WebFetch upstream → claude-code-guide → deep-research; match
the PINNED version, not "latest".
- Graceful degradation: the plugin accelerators may be absent on a fresh checkout,
so the policy commits to the principle and falls back to WebFetch/WebSearch.
CLAUDE.md gains a concise operating rule + pointer. TODO 10.4 decided. New TODO
10.7 tracks making the plugin/MCP toolchain reproducible from the repo (it is not
synced by account nor carried by git) — surfaced by a portability check this run.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Resolve the conflict between ADR-011 (tags-not-digests) and the security work
(digest pinning) with one coherent rule that respects ADR-011's stateless/stateful
split:
- Stateful → pin `tag@digest` (readable tag + integrity digest): legible diffs AND
tamper-evidence. Snapshots cover broken updates; the digest covers swapped images.
- Stateless → rolling tags (latest/stable); digest-pinning would defeat the rolling
design. Integrity rests on official/verified images + disposability.
Aligned across ADR-011 (decision 2), ADR-004 (image management), ADR-002
(supply-chain row), accepted-risk R1, the service checklist, and TODO 15.6.
TODO 16.7 marked decided.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
ADR-013 sets how boma draws on AnsibleBaobabV4 without inheriting it:
translate-don't-transplant — V4 is evidence, never authority. It is a legitimate
source only of operational gotchas and working config snippets (re-derived on
boma's terms); never requirements, domain values, structure, or conventions.
Provenance stays transient (commits/conversation), durable docs stay clean. AI
consultation guardrails included. Resolves TODO 3.3 and 10.1.
Also bring ADR-011 (update management, Proposed draft) under version control:
- fix its "reuse V4's ntfy topics" line to "boma defines its own" (ADR-013)
- track its 6 open questions in TODO 16, plus a 7th: reconcile its tags-not-digests
pinning with the digest-pinning the security work now mandates (R1 / checklist /
15.6) — they currently conflict.
CLAUDE.md gains a V4 guardrail + ADR-013 pointer.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Revise ADR-004 to a service-role standard: every service is its own
self-contained role with a required file set including SECURITY.md, uniform
deploy mechanics, and a deferred shared-engine option (with revisit trigger)
recorded in the ADR.
Add the per-service security record:
- docs/security/service-security-template.md — canonical SECURITY.md template
(exposure, checklist status, service-specific hardening, residual risks)
- roles/<service>/SECURITY.md is where each service records how it meets the bar;
/security-review aggregates roles/*/SECURITY.md and cross-checks against config
- service-checklist.md noted as the generic bar the record answers
Wire-up: new-role runbook step writes SECURITY.md from the template; ADR-002
governance bullet points at it; CLAUDE.md role conventions require it and mandate
one-role-per-service; STATUS records the convention as defined-not-yet-applied.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Walked the seeded accepted-risk register (R1-R4) and turned inherited gaps into
deliberate decisions:
- Supply chain (R1): tightened to required baseline hygiene (digest pinning,
official/verified images); active scanning deferred — stays an accepted risk
- CIS (R2): adopted as a positive decision — CIS Debian L1+L2 (base role) + CIS
Docker (docker_host + service checklist); app layer via the checklist
- SELinux/AppArmor (R3): AppArmor becomes a baseline control (CIS-enforced);
register keeps a clean "no SELinux" accept
- IDS (R4): adopt AIDE (baseline via CIS) + Suricata on OPNsense + active alerting
Register shrinks from 4 inherited gaps to 2 deliberate accepts. ADR-002 gains a
Hardening standard section; STATUS + TODO 15 track the (unbuilt) implementation,
including the CIS L2 partition impact on VM provisioning (ADR-006).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add a managerial security frame on top of the host baseline: explicit threat
model (opportunistic external, lateral movement/blast radius, operator/agent
error; supply chain accepted-lower-priority), security principles, and four
governance mechanisms that ADR-002 establishes and links out to:
- docs/security/service-checklist.md — per-service security bar (referenced
from the new-role runbook)
- docs/security/accepted-risks.md — living accepted-risk register (R1-R4)
- planned /security-review skill (TODO 8.5)
- agent guardrails in CLAUDE.md "what Claude must not do"
STATUS.md records the frame as present (manual enforcement) and /security-review
as planned-not-built.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Forgejo origin is trunk-based with no merge-request gate, so the
finishing-a-development-branch "open a PR" option doesn't apply — merge
locally then push. Also carries earlier uncommitted FRICTION.md edits
(emphasis normalization + 2026-05-31 ADR-status entry).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Final-review minor: the /capacity-review skill overwrites a latest.md
pointer alongside the dated report; record that in the ADR.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Brainstormed design for docs/hardware/reference.md (physical compute +
network gear + workload placement intent), a stdlib-only capacity-scan.py,
and an on-demand /capacity-review skill that reports to docs/hardware/reviews/.
Mirrors the repo-scan -> /review-repo -> docs/reviews triad.
TODO additions: schedule /capacity-review later and decide its usage-stats
source (Proxmox RRD vs the Prometheus/Loki/Grafana/Alloy stack) before
building any hook (8.4); reevaluate the stdlib-only script policy (#14).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
docs/FRICTION.md: a running log of friction/gotchas/recurring-fixes/unused tooling,
seeded with this session's real signals — raw material for the periodic kaizen
review. docs/TODO.md: schedule building /retro in ~1 week, and record the Claude-setup
decision. (Also carries your earlier backlog edits.)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
ADR-010: API tokens as least-privilege managed secrets, declarative-first (no
click-ops), automation boundary, planned trunk-based CI. CLAUDE.md/AGENTS.md:
check 'rbw unlocked' before vault-dependent tasks (incl. commits) rather than
failing partway.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Forgejo's /raw/ API is read-only so it cannot serve as a Terraform HTTP state
backend. Switch both envs to local state on the control node (ADR-006); remove
the dead TF_HTTP_* credential hints.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
R6/R7: ADR-003 & ADR-008 CI pipelines rewritten trunk-based (push to main ->
test -> staging -> [manual gate] production); CLAUDE.md no longer forbids pushing
to main. R8: STATUS/roles-README/site.yml now say base & docker_host are not built
(not in git), so a clean clone errors. R15/R16: ADR-001 table flagged as intended
design; dropped the unbuilt 'monitoring agent' from the baseline.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Point ADR-005, the new-host runbook, CONTRIBUTING, and AGENTS at the
rbw/Vaultwarden flow instead of a .vault_pass file. Also record the cron-section
idea in docs/TODO.md.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
First /review-repo run on boma. Hardened repo-scan.py (no TODO.md/prose false
positives). Applied 7 safe fixes (DNS staleness x2, STATUS factual correction,
hosts.yml path generalisation, trunk-based wording x2, scripts/README). Recorded
the run and 17 open findings in docs/reviews/2026-05-30-*.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
