4 Commits

Author	SHA1	Message	Date
claude-ceo-assistant (Claude Opus 4.7 on Hongming's MacBook)	16868c4ec1	fix(plugins): SaaS (EC2-per-workspace) install/uninstall via EIC SSH ... Closes the 🔴 docker-only row in docs/architecture/backends.md. Plugin install on every SaaS tenant currently 503s with "workspace container not running" because the handler is hardcoded to Docker exec but SaaS workspaces live on per-workspace EC2s. Caught on hongming.moleculesai.app when canvas POST /workspaces/<id>/plugins surfaced the error. Mirrors the Files API PR #1702 pattern: dispatch on workspaces.instance_id in deliverToContainer (and Uninstall). When set, push the staged plugin tarball to the EC2 over the existing withEICTunnel primitive (template_files_eic.go) and unpack into the runtime's bind-mounted config dir (/configs for claude-code, /home/ubuntu/.hermes for hermes — see workspaceFilePathPrefix). chown 1000:1000 to match the docker path's agent-uid contract; restart via the existing dispatcher. Direct host write rather than docker-cp via SSH because the runtime's config dir is already bind-mounted into the workspace container — the runtime sees the files on next start with no additional plumbing. Adds InstanceIDLookup (parallel to RuntimeLookup) so unit tests don't need a DB; production wires it in router.go like templates.go does. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-07 15:42:51 -07:00
Hongming Wang	62fc25757c	docs(backends): document Auto-dispatcher SoT pattern + source-level pins ... Closes #10. The 2026-05-05 hongming silent-drop incident shipped because the backends.md parity matrix didn't enforce a "go through the dispatcher" rule — three handlers (TeamHandler.Expand, OrgHandler.createWorkspaceTree, workspace_crud.go's stopAndRemove) silently bypassed routing on SaaS for ~6 months across two distinct verbs. This doc pass: - Adds a "How to dispatch" section that's the canonical answer to "where do I call Start / Stop / Has from?". Names the three dispatchers (provisionWorkspaceAuto, StopWorkspaceAuto, HasProvisioner), their fallbacks, and the allowed exceptions. - Updates the matrix lifecycle rows so every dispatched operation points at the dispatcher source, not the per-backend bodies. - Adds Org-import + Team-collapse rows so the bulk paths are visible to anyone scanning for parity gaps. - Lists the source-level pins (4 of them) under Enforcement so future contributors see them as load-bearing tests, not noise. - Adds a "When you add a NEW dispatch site" section so the next verb (Pause / Hibernate / Snapshot) lands as a dispatcher mirror, not as another bespoke handler that drifts from the existing two. - Refreshes Last audit to 2026-05-05. No code change; doc-only. The SoT abstractions described here landed in PRs #2811 + #2824. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-04 20:25:10 -07:00
Hongming Wang	b54968878a	docs(internal): refresh runtime-package mirror policy + parity matrix + dead-link fix ... - workspace-runtime-package.md: add explicit "Where to make changes" section documenting the mirror-only policy on Molecule-AI/molecule-ai-workspace-runtime — direct PRs are auto-rejected by mirror-guard CI; staging push regenerates both the mirror and the PyPI wheel via .github/workflows/publish-runtime.yml. - infra/workspace-terminal.md: replace dead molecule-core#1528 reference (repo renamed to molecule-monorepo, no longer accepting issues at the old name) with a forward-pointer to monorepo + molecule-controlplane issue trackers. - architecture/backends.md: bump audit date to 2026-05-02 and add rows for channel envelope enrichment (#2471), chat_history MCP tool (#2474), /activity before_ts paging (#2476), /activity peer_id filter (#2472), runtime_wedge smoke gate (#2473 + #2475), and the canvas-E2E state-file requirement (#2327). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-01 20:06:06 -07:00
Hongming Wang	a56b765b2d	docs: testing strategy + PR hygiene + backend parity matrix + boot-event postmortem (#1824) ... Bundles the documentation and lightweight tooling landed during the 2026-04-23 ops/triage session. Pure additions — no behavior changes. ## Added ### docs/architecture/backends.md Parity matrix for Docker vs EC2 (SaaS) workspace backends. 18 features tabulated with current status; 6 ranked drift risks; enforcement hooks (parity-lint + contract tests). Living document — owners are workspace-server + controlplane teams. ### docs/engineering/testing-strategy.md Tiered test-coverage floors instead of a blanket 100% target. Seven tiers by code class (auth/crypto → generated DTOs). Per-package current-state snapshot + targets. Tracks the 3 biggest coverage gaps (tokens.go 0%, workspace_provision.go 0%, wsauth ~48%) against their tier-1/2 floors. ### docs/engineering/pr-hygiene.md Captures the patterns that keep diffs reviewable. Motivated by the 2026-04-23 backlog audit where 8 of 23 open PRs had 70-380-file bloat from stale branch drift. Covers: small-PR sizing, rebase-not-merge, cherry-pick-onto-fresh-base for recovery, targeting staging first, describing why-not-what. ### docs/engineering/postmortem-2026-04-23-boot-event-401.md Postmortem for the /cp/tenants/boot-event 401 race. Root cause (DB INSERT ordered AFTER readiness check), detection path (E2E + manual log inspection), lessons (write-before-read pattern, integration tests needed, E2E alerting gap, invariants-as-comments). ### tools/check-template-parity.sh CI lint for template repos — diffs the `${VAR:+VAR=${VAR}}` provider- key forwarders between install.sh (bare-host / EC2 path) and start.sh (Docker path). Catches the #5 drift risk from backends.md before it ships. ### workspace-server/internal/provisioner/backend_contract_test.go Shared behavioral contract scaffold for Provisioner + CPProvisioner. Compile-time assertions catch method-signature drift today; scenario- level runs are t.Skip'd pending backend nil-hardening (drift risk #6, see backends.md). ## Updated ### README.md Links the new engineering docs + backends parity matrix into the Documentation Map so agents and humans can actually find them. ## Related issues - #1814 — unblock workspace_provision_test.go (broadcaster interface) - #1813 — nil-client panic hardening (drift risk #6) - #1815 — Canvas vitest coverage instrumentation - #1816 — tokens.go 0% → 85% - #1817 — 5 sqlmock column-drift failures - #1818 — Python pytest-cov setup - #1819 — wsauth middleware coverage gap - #1821 — tiered coverage policy (meta) - #1822 — backend parity drift tracker Co-authored-by: Hongming Wang <hongmingwang.rabbit@users.noreply.github.com> Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> Co-authored-by: molecule-ai[bot] <276602405+molecule-ai[bot]@users.noreply.github.com>	2026-04-23 19:59:38 +00:00