Fixed two concrete gaps in the MCP knowledge-search endpoint: no locale field
existed at all (now nl-BE/en-GB/fr-BE, wired to the knowledge provider's
existing language param), and the correlation ID was always freshly minted,
ignoring any inbound X-Correlation-Id header. Added a shared dependency and
applied it to all four MCP endpoints so Fleet Ops's own audit log preserves
the Hub's real correlation ID end to end.
MCP_HUB_BASE_URL/MCP_PROVIDER_ID were declared in .env.example but never read
anywhere. Since the Hub's own registration is catalog-driven (it never needs
Fleet Ops to push a registration call), wired mcp_hub_base_url for a real Hub
reachability health check instead of an unneeded self-registration call.
Renamed Fleet Ops's own internal audit tool labels mobilityops_* -> fleet_ops_*
(mirrored in contracts/mcp-tools.json with mobilityops_* kept as deprecated
aliases); documented that the live Hub connector's own dotted tool namespace
is a separate, Hub-owned naming layer, deliberately not touched.
Automation page's MCP card now shows real evidence (last tool/client/count/
timestamp, honest no-evidence state) instead of just the registration flag.
Also fixed a real methodology gap found mid-session: compose.yaml's api
service has no bind mount, so `docker compose run --rm api` silently tests a
stale image until rebuilt. Re-ran every local gate after rebuilding; fixed one
genuinely stale test assertion and two lint line-length errors surfaced by
that rebuild. 176 tests passing, ruff clean, mypy clean (50 files).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Pushed feat/live-n8n-ragcore-integration to origin, then redeployed
the live Fleet Ops instance from 0da5251 to 0571a40 following the
deployment directory's own established source-archive convention.
Verified live: /health OK, the new n8n procedures endpoint (added
this branch) is reachable and correctly auth-gated, KNOWLEDGE_PROVIDER
still demo as intended.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Root cause found via the live n8n container's own logs: N8N_PROXY_HOPS=0
in the Unraid template didn't match the real reverse-proxy in front of
it, breaking the browserId/CSRF check on every workflow save while
leaving the UI looking fully signed in -- the user's pushback that it
"shows logged in" was correct and prompted digging into server logs
instead of continuing to guess client-side.
Fixed by editing the Unraid template (N8N_PROXY_HOPS 0->1, backed up
first) and recreating the container with every other setting preserved
exactly. Verified by reproducing the exact save action that used to
fail; it now works, and node persistence survives a full reload.
Built and saved both remaining workflow-3 nodes (Summarize sync
result, Report sync result to Fleet Ops) after recovering from an
errant Ctrl+A that deleted a node mid-verification (caught via node
count, restored via n8n's own Version History, redone carefully). Not
published -- that starts real daily production runs and is left for a
separate decision.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Issued a fresh scoped credential via the RAGcore admin UI (separate,
working OIDC session, unaffected by the n8n auth problem) and made a
real authenticated /v1/answers call. Got a clean 200 with a real
answer_id/retrieval_run_id, not degraded -- but not_answerable, 0
citations.
Confirmed this isn't a regression: the same query through RAGcore's
own pre-existing Query Lab tool (untouched this session) returns the
identical result down to zero dense/sparse candidates at the raw
retrieval stage. Ruled out the obvious causes via direct Qdrant/
Postgres checks -- workspace_id, space_id, status, and embedding
digest all correctly match real indexed content. Root cause not yet
found; flagged as a follow-up rather than pursued further to avoid
scope creep on what was a deployment-verification task.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Deployed a2905cc to the live RAGcore instance (approved). Verified
the fix directly: POST /v1/search now returns 401
AUTHENTICATION_REQUIRED instead of the old permanent 503
SEARCH_UNAVAILABLE, proving the endpoint reaches real request
handling. A full authenticated /v1/answers call with a real grounded
answer is still outstanding -- the previously-issued production
credential's raw token was never persisted anywhere retrievable, and
issuing a fresh one needs the RAGcore admin UI, not attempted this
round.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The live "Fleet Ops -- RAGcore Procedure Sync" workflow's canvas was
found at zero nodes -- the earlier session's abandoned direct n8n
REST API attempt had gone far enough to wipe it before hitting its
401. Recovered via n8n's own Version History "Restore version"
action back to the last good 4-node save; verified via DOM node
count before and after.
Adding the two remaining nodes then hit the same failure mode: n8n's
own first-party autosave reported "Unauthorized" moments after a
fresh, successful interactive sign-in. Stopped deliberately rather
than retrying against a live instance that already caused one data
loss incident this session -- this looks like an n8n-side session/auth
problem, not something fixable from browser automation.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The previous adapter targeted an endpoint shape RAGcore never actually
exposed. health() now checks /health/ready and ask() posts to the real
POST /v1/answers with Bearer auth and requested_space_ids, matching
RAGcore's actual contract after this session's Bearer-auth and
search/answer wiring work.
Adds RAGCORE_SPACE_ID config/env plumbing (a question is meaningless
without a knowledge space to scope it to) and 12 new adapter tests
covering degradation paths: missing space id, connection errors,
non-200 responses, malformed responses, not-answerable, and
answerable-without-citations all fail closed to "insufficient
evidence" rather than fabricating an answer.
KNOWLEDGE_PROVIDER stays "demo" in production for now -- switching
requires RAGcore's own search/answer application to actually be
deployed and live-verified, tracked separately in PROJECT_STATE.md.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
With explicit owner approval, traced the persistent credential-issuance
rejection to a cross-transaction race in RAGcore's own dependency
injection (two independent DB transactions per request instead of one
shared transaction), fixed and deployed it in RAGcore, and verified a
working "RAGcore Sync Token" n8n credential now exists. Unblocks
workflow 3 and the RAGcoreKnowledgeProvider adapter rewrite.
User explicitly authorized issuing the RAGcore credential directly this
round. Retried via the admin UI (Platform Admin role) after the earlier
raw-API attempt; both fail with an opaque server-side rejection carrying
a trace ID. Documents this as a RAGcore-side blocker, not a Fleet Ops gap.
New central "Fleet Ops — Workflow Error Handler" workflow (Error
Trigger -> safe-report Code node -> POST to the new /workflow-error
endpoint), wired as the Error Workflow on both existing workflows with
no recursive loop on itself. Live-validated end-to-end against the
real Fleet Ops server (register + idempotent re-register), and via a
genuine induced failure on the scheduled-scan workflow (broken URL,
confirmed failure, reverted, confirmed healthy).
Fixed two real bugs found during live testing: Code node needed
"Run Once for Each Item" (not "All Items") for $json binding, and
every HTTP body field had a stray trailing space from the n8n
code-editor's bracket auto-close that broke datetime/enum validation.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Move the two live-validated workflows into n8n/workflows/ (credential-
based auth referenced by name only, no secret values), add a manifest
covering all 4 canonical workflows and a read-only drift-check script
against n8n's Public API. Retire the pre-integration root-level starter
files that still carried the literal-token pattern, and repoint the
Unraid deploy scripts, Makefile targets and runbook at the new files.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
PROJECT_STATE.md: fix the stale "Product name: MobilityOps."/"PoC only"
locked-decisions lines (predate the Fleet Ops rebrand), fix the "Fleet
Ops correction" section header still reading "IN PROGRESS .../Not yet
merged to master" when it was in fact already merged (de0bdea, evidence
commit f780557), and append a new dated entry for this correction round
with commits and gate evidence so far.
README.md: reference docs/fleet-ops-final-localization/ alongside the
existing docs/fleet-ops-correction/ link, refresh the stale Playwright
test count (113 -> 138).
Deployed fix/fleet-ops-i18n-status-flow to http://192.168.10.150:1236 and validated
live, which directly caught the evidence-summary localization bug (fixed in 2e4fb43).
Redeployed with the fix and re-verified: full 116-test Playwright suite green against
the live server, no console errors, no container-log errors, both containers healthy,
scenario_integrity.all_ready: true after final reset.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Fresh clone of only committed files into an isolated Compose project (separate ports,
no shared volumes) validated: migration from empty database to head, deterministic
seed (matches the corrected 27-issue count), 151 backend tests + Ruff + mypy, frontend
build, and the full 113-test Playwright suite -- all green. Isolated stack torn down
afterward; working dev environment confirmed untouched. Full detail in
PROJECT_STATE.md; test counts refreshed in README.md.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
- contracts/openapi.yaml: title is now "Fleet Ops API"; documents the new
status-recommendation preview endpoint and the apply endpoint's request body
(recommendation_token) and full error-code set; notes the search endpoint's
code+params response shape.
- README.md: title and intro now say Fleet Ops, with an explicit note on the
Fleet Ops (visible)/MobilityOps (technical identifier) naming split; refreshed
stale test counts (151 backend, 108 Playwright).
- PROJECT_STATE.md: full progress record for the in-progress correction milestone,
including what's done, what bugs were found and fixed, and what's explicitly not
yet done (i18n test-strengthening 11D/E/F, clean-checkout drill, Unraid deployment,
merge to master).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Adds artifacts/fleet-ops-release/final-summary.md with the complete evidence trail for
this release: commits, branding, locale/translation/knowledge-base coverage, adaptive
Demo Guide behaviour per breakpoint, Data Quality/Automation/Audit/clickable-row
improvements, full test results (backend, lint, build, 92 Playwright tests) re-run
against the local stack, an isolated clean-checkout drill, and both the feature-branch
and post-merge master deployments to Unraid -- plus 10 screenshots across the three
languages, desktop and mobile. Updates PROJECT_STATE.md with the corresponding summary.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Records the clean-checkout drill result (127 backend tests, 56
Playwright tests, all green on an isolated fresh clone), final live
Unraid verification, and the complete required evidence summary for
the demo-productization work on this branch.
Adds one comprehensive Playwright test that walks a fresh Operations
Manager session through all 8 Demo Guide steps performing the real
action at each one, then restores the environment. Writing it surfaced
a real desktop layout bug: the Demo Guide's fixed side panel overlapped
main content with no reflow, making the return form's "Review return"
button unclickable while the guide was open at ordinary viewport widths.
Fixed by reserving layout space via a guide-open class. Also adds mobile
bottom-sheet, keyboard-reachability, and console-error checks.
Integration status badges across Dashboard/Automation now show honest
plain-language labels instead of raw backend state strings (and fix a
few states that had no matching CSS colour class at all). Audit trail
gets a "view related events" action reusing the existing correlation_id
filter. About page gains scope/architecture/security/testing sections
and a guided-demo entry point. POST /api/v1/demo/reset now runs and
records a server-side scenario-integrity check. Also fixes a second real
race condition (caught by the return-review e2e test): the odometer
scenario pre-fill now resolves before ReturnForm mounts instead of
patching its value in after the fact.
Fixes a real honesty bug in Knowledge.tsx (body copy named "RAGcore" while
the active provider is the demo one) and a second real bug discovered
while fixing it: the brief's suggested Dutch questions would silently
return "insufficient evidence" against the English-only demo knowledge
base -- verified empirically and fixed by keeping suggested questions in
English. The return flow now pre-fills the odometer-regression scenario's
suspicious reading instead of asking a visitor to invent one, and links
to automation/audit after committing. Data-quality issues get a shared
plain-language "what's wrong / why it matters" explainer per rule type,
a post-resolution confirmation with audit/vehicle links, and a "demo
scenario's only" list filter. Also fixes a real async race where the
odometer pre-fill could clobber text a visitor had already started typing.
Adds a compact "Probeer een demonstratiescenario" page listing the 5 named
scenarios with live readiness from the manifest, plus a Demo Guide side
panel (bottom sheet on mobile) that walks an Operations Manager through
all 8 steps with per-step context, live-resolved routes, sessionStorage
progress, and a "Demo opnieuw voorbereiden" restart. Login's guided-demo
CTA now actually opens the guide. Fixes a real mobile topbar overflow the
new guide trigger introduced.
Adds GET /api/v1/demo/manifest as a single source of truth for the demo's
fictional org identity (Northstar Mobility -- surfacing the project's
already-locked tenant name), synthetic-data/reset state, and live scenario
readiness. Rewrites the login screen in Dutch with an honest, no-password
demo entry and a guided-demo entry point, replaces the loud full-width
demo banner with a subtle badge + popover, and adds a compact About page
explaining what's real vs. synthetic vs. not yet connected.
Booking/inspection/maintenance/outbox dates were authored as absolute
timestamps around a fixed 2026-08-01 anchor and never re-anchored at
seed/reset time, so demo scenarios (e.g. BK-DEMO-RETURN) silently drifted
into the past. Every reset now shifts seeded dates by (today - authored
anchor); dashboard's "today" filter uses real wall-clock time instead of
the now-removed frozen demo_today setting. Adds seed-validation tests
proving scenarios S1/S2/S4/S5 are present and internally consistent after
every reset.
Ran a dedicated post-M7 release-readiness audit. Found and fixed the one real gap: mypy
was a declared dev dependency but had never been run in any milestone's validation loop.
Fixed all 43 pre-existing type errors it surfaced, including two genuine defensive-
programming gaps (unguarded Optional vehicle/customer lookups that could have crashed
with unhandled 500s instead of clean 404/401 responses) rather than suppressing them.
make lint now runs ruff + mypy; mypy reports zero errors across 44 source files.
Re-verified end to end against a genuinely wiped-volumes clean checkout: automatic
migrations, deterministic seed, 66/66 backend tests, and the full user-journey matrix
(login, dashboard, vehicle/booking detail, return workflow, invalid-mileage rejection,
data-quality review, duplicate-customer merge, audit trail, Knowledge Assistant, n8n,
MCP Hub) via curl and Playwright.
Live-verified both external-dependency degraded modes, not just unit tests: stopped n8n
mid-flow and confirmed a return still commits with the outbox event staying pending and
retrying with backoff, then self-healing to succeeded with zero manual intervention once
n8n came back; verified RAGcore's unavailable-degradation path against an unreachable
host. Added frontend/e2e/interactive-elements.spec.ts (11 tests covering every nav item,
filter, tab, and role boundary) alongside the existing demo script test — 12/12 e2e tests
passing.
Verified no secrets are committed (.env never tracked, clean git history scan) and
.env.example covers every operator-configurable setting. Confirmed no placeholders,
TODOs, fake responses, hardcoded metrics, or dead routes anywhere in the codebase.
Updated README.md with an honest integration-status section and PROJECT_STATE.md with
the full audit findings. Added artifacts/final-acceptance/summary.md as the authoritative
final evidence document (commands, results, URLs, demo access, integration status per
external dependency, known limitations, deployment instructions, five-minute demo flow).
Automated migrations on container startup (backend/entrypoint.sh), scripted n8n workflow activation (make n8n-setup), Playwright E2E test covering the full 9-step demo script (verified passing against the live stack, including the previously-unverified 360px responsive layout), evidence screenshots of all main pages, architecture diagram, and artifacts/evidence/final-summary.md with commit/commands/test counts/RAGcore and n8n evidence/MCP sample calls/known limitations/portfolio wording. Verified the complete clean-checkout path from a genuinely wiped-volumes state: automatic migrations, seed, 66 backend tests passing, and a live S1 return round-tripped through a freshly-activated n8n instance. Added .gitattributes to force LF line endings on shell scripts, preventing a real cross-platform breakage of entrypoint.sh's shebang.
Four read-only, service-token-protected MCP provider endpoints (operations summary, attention vehicles, vehicle details, knowledge search facade). Shared-secret auth reusing the M4 callback pattern. Service-request audit trail for every call. Extracted shared operations-summary logic out of the dashboard router to avoid duplicating retrieval logic. 66 backend tests passing, ruff clean. Verified all four endpoints and audit trail directly via curl against the live stack (no live MCP Hub instance available in this environment).
KnowledgeProvider protocol with a deterministic TF-IDF-weighted extractive demo provider (never generative, always cites real excerpts) and a RAGcore HTTP adapter that degrades cleanly to unavailable. Knowledge nav + chat-style Q&A UI with source cards and honest grounded/insufficient/unavailable states. 57 backend tests passing, ruff clean. Fixed a real relevance bug (generic terms like "vehicle" crowding out distinctive matches) via IDF weighting, found by testing the actual S6 scenario. Verified end-to-end in the browser: grounded damage question cites both expected procedures; unrelated question honestly returns insufficient evidence with no fabrication.
Outbox dispatcher (background thread, FOR UPDATE SKIP LOCKED claim, exponential backoff, no transaction held during HTTP I/O). n8n callback endpoint with shared-secret auth and idempotency by event ID. Automation nav + UI with manual retry. 49 backend tests passing, ruff clean. Fixed a crash-on-redelivery bug in seeded outbox payloads and made the dispatcher defensive against malformed payloads. Verified the full live round trip against a real n8n instance: return -> outbox -> dispatcher -> n8n workflow -> callback -> succeeded, including the S5 failed-retry demo scenario.
Five rule scanners (duplicate customers, missing fields, odometer regression, booking overlap, status conflict) run automatically after seed and via an explicit scan endpoint. Issue defer/reject/merge-customers endpoints with transactional customer merge (booking rewiring, tombstone, audit). Data Quality nav + workbench UI with two-column duplicate comparison and inline (non-native) confirm. Dashboard attention items now link to issues. 35 backend tests passing, ruff clean. Fixed a real false-positive bug in odometer-regression detection found through iteration on seed data, and two TS narrowing errors. Verified end-to-end via browser: S2 merge and S4 overlap scenarios.
