Adds regression tests for the fixed-width timestamp format and sub-second
boundaries, in-memory store dedup and cursor pagination, stream-ID selector
escaping, live-vs-synthetic event IDs, gap healing and baseline state reset,
path-scoped diff ignores, nested Added/Removed flags, snapshot watermarks and
the TypeSnapshot record, NodePort/headless service sanitization, the
dry-run-gated apply, skipped-object reporting, and the expanded gvrFor table.
The object-history endpoint now scopes gap markers to the requested window,
and record --server fails loudly instead of silently recording locally.
Audit: correlation event_ids include stage and response code so the
multi-stage lines of one request no longer collapse under dedup; match
scans are bounded to a time window and a page instead of the object's full
history.
Web: dry-run results are read from dry_run_result (nested) so the verdict is
rendered correctly and conflicts/errors/skipped are shown; the plans view no
longer POSTs an unsolicited plan on page load; coverage and streams surface
API errors instead of showing a misleading empty state and follow cursor
pagination so they are not stuck on the oldest page; the diff path tokenizer
handles backslash-escaped dotted keys.
Deploy/CI: the chart no longer grants the query server a cluster-wide read
ClusterRole, runs as non-root with a read-only root filesystem, adds
liveness/readiness probes, wires the ConfigMap as env (STORE_PATH,
LISTEN_ADDR), defaults the journal to a PVC instead of an ephemeral
emptyDir, and adds imagePullSecrets; the replay ClusterRole drops the
unused update verb; a Dockerfile builds a static distroless image; GitHub
actions are pinned by commit SHA, jobs set least-privilege permissions, the
Vercel deploy skips fork PRs, CI passes the Makefile test timeouts, and
make lint runs a real web syntax check.
Docs: event-schema/consistency no longer describe an ingest_sequence field,
an observed-time-based event_id, or restart-from-checkpoint; the threat
model documents the unauthenticated HTTP API surface and the chart's RBAC
change; replay-safety matches the enforced dry-run gate.
Storage: fixed-width millisecond timestamp format with a one-time migration
so lexicographic comparisons stay chronological; snapshot auto-IDs get
nanosecond precision; Close is synchronized; commit failures roll back; DSN
paths are URI-escaped; the dedup index is no longer rebuilt every startup;
the in-memory store now dedups by event_id and honors Limit/Offset/SinceSeq.
Event/discovery: stream IDs URL-escape the selector so slash-containing label
selectors round-trip and cannot collide; alias resolution fills empty fields
only and no longer clobbers explicit group/version/kind; rc maps to
replicationcontrollers and rs to replicasets.
Watch: the label selector is now actually applied to List and Watch; the 410
relist path is backoff-throttled, backoff resets after a healthy watch, an
empty list resourceVersion backs off instead of looping forever, non-410
watch errors reconnect without writing a spurious permanent gap, an idle
watch timer forces reconnects, and Run cancels sibling streams and recovers
from panics. Synthetic relist events carry a distinct event_id so unchanged
objects re-listed after a gap survive dedup.
Materialize: a baseline now resets object state (objects deleted during a gap
disappear) and heals open gaps; diff ignores server metadata by path, not by
key name, so user fields named status/uid are kept; nested add/remove carry
Added/Removed flags; dotted paths are escaped; Diff validates the window and
treats a zero before as empty; Snapshot records per-stream watermarks and
writes a TypeSnapshot journal record.
Replay: apply requires a successful dry run; gvrFor covers every policy kind
so Include* toggles work; unsupported kinds are reported as skipped instead
of silently dropped; NodePort services lose clusterIP and nodePort while
headless keeps clusterIP: None; targetNS is always honored; namespace mapping
collisions are detected; plan IDs and field managers are collision-safe and
plans are mutex-guarded against concurrent dry-runs/applies.
API: cursor pagination pushes ingest_seq into the SQL filter (export --server
no longer truncates); gap records expose their payload; /v1/diff requires a
cluster_id; dry-run/apply no longer accept a client-supplied kubeconfig;
plans carry target_context; HTTP server gets timeouts.
Metrics: the watch collector and replay planner now bump the registered
counters, and the server refreshes store-derived gauges on a ticker.
CLI: replay apply works locally and refuses to apply after a failed dry run;
coverage --server follows cursors and shows gap details; timeline
auto-detects namespaces and resolves cluster-scoped objects; diff validates
the window; export writes 0600 files; PVCs are no longer treated as
cluster-scoped; krply-server reports the build version.
- fall back from in-cluster to KUBECONFIG and ~/.kube/config so a
cluster-less host reports a clear error instead of a raw token path
- regression test for explicit kubeconfig loading
- wrap all routes with CORS, honoring KRPLY_CORS_ORIGINS (default *)
- short-circuit preflight OPTIONS and echo allowed origins
- honor the PORT env var for platform deployments (render etc)
- grpc query server for events, snapshots, diffs, and replay plans
- server metrics, request audit logging, and identity endpoint
- krply-server and krply cli entrypoints
- dependency-light dark web ui backed by the query api
- plan replay paths across the journal with irreversible-deletion and apply-order checks
- dry-run and live apply via list-and-watch with backoff
- produce human-readable path summaries for each resource
- durable event schema with stream identity and dedup hashing
- SQLite WAL and in-memory journal stores with atomic append+checkpoint
- list-and-watch collector with bookmarks, reconnects, and 410 relists
- state reduction, snapshots, and semantic field diffs
