fix: harden gossip, HLC restarts, watcher races, MCP args, pagination (#3)

Implements the top findings from the codebase review, verified with tests and live CLI/MCP checks.

**Gossip integrity**
- Push validation: 4 MiB body cap, 1000-row batch cap; rows claiming the local node id (vector-poisoning), empty node ids, and negative HCLs rejected (internal/watch/gossip.go, internal/db/gossip.go)
- Reconcile-on-pull: Run returns the pulled count, syncGossip rebuilds derived state when > 0 — entry-count comparison could never fire, so synced observations never materialized into searchable entries

**Data-layer safety**
- HLC resumed from MAX(hcl) at Open (hlc.SeekTo): a restart with a regressed wall clock cannot reissue values the (node_id, hcl) locator and pull cursors depend on
- Writer serialization: _txlock=immediate DSN + SetMaxOpenConns(1) + per-KnoxDB mutex around RecordObservation's check-then-insert dedup (closes duplicate-row race)

**Watch daemon**
- Ticker guard flags now atomic.Bool (was a cross-goroutine data race)
- Trailing-edge per-path debounce (timer-based, pruned on fire/delete)
- Recursive watches (startup tree walk + watcher.Add on dir Create); Rename re-ingests, Remove cancels pending ingests

**MCP + CLI**
- Strict arg validation, no silent clamping: thread_id 0 errors instead of renaming thread #1; empty knox_thread_link {} errors instead of false success; thread existence checked before writes; golden-thread tool nil-safe
- --page 0 errors instead of panicking; query/recent pagination actually pages (page x limit)

**Tests** (new internal/hlc and internal/db packages): SeekTo monotonicity, concurrent dedup race, push validation, reopen HCL monotonicity, batch caps, self-spoof rejection, idempotency on observation counts.

Verified: go build, go vet, full suite with -race, live MCP stdio transcripts against a scratch DB.
Reviewed-on: #3
Co-authored-by: David Gwilliam <dhgwilliam@gmail.com>
Co-committed-by: David Gwilliam <dhgwilliam@gmail.com>
This commit was merged in pull request #3.
This commit is contained in:
2026-09-17 09:06:08 +00:00
committed by david
parent d6d2a24ddc
commit bb852faa27
10 changed files with 733 additions and 116 deletions
+19 -6
View File
@@ -13,7 +13,7 @@ import (
func paginate(entries []db.Entry, page, limit int) []db.Entry {
start := (page - 1) * limit
if start >= len(entries) {
if start < 0 || start >= len(entries) {
return nil
}
end := start + limit
@@ -74,16 +74,22 @@ func NewQueryCmd(kdb *db.KnoxDB) *cobra.Command {
if query == "" {
return fmt.Errorf("search query required (positional arg or stdin pipe)")
}
results, err := kdb.Search(query, limit)
if page < 1 || limit < 1 {
return fmt.Errorf("page and limit must be >= 1 (got page=%d limit=%d)", page, limit)
}
// Fetch enough rows to cover the requested page: Search caps at its
// limit argument, so slicing a limit-sized result set could never
// reach page 2+.
all, err := kdb.Search(query, page*limit)
if err != nil {
return err
}
results = paginate(results, page, limit)
results := paginate(all, page, limit)
if len(results) == 0 {
fmt.Println("No results found.")
return nil
}
fmt.Printf("Found %d results for %q (page %d, %d per page):\n\n", len(results), query, page, limit)
fmt.Printf("Found %d results for %q — showing %d (page %d, %d per page):\n\n", len(all), query, len(results), page, limit)
for _, r := range results {
fmt.Printf(" %-8s %-30s [%s] %s\n", shortFP(r.Fingerprint), truncateStr(r.Title, 30), r.Project, r.SourceID)
if r.Summary != "" {
@@ -105,13 +111,20 @@ func NewRecentCmd(kdb *db.KnoxDB) *cobra.Command {
Use: "recent",
Short: "Show recent knowledge entries",
RunE: func(c *cobra.Command, args []string) error {
all, err := kdb.RecentEntries(limit * 10)
if page < 1 || limit < 1 {
return fmt.Errorf("page and limit must be >= 1 (got page=%d limit=%d)", page, limit)
}
all, err := kdb.RecentEntries(page * limit)
if err != nil {
return err
}
entries := paginate(all, page, limit)
if len(entries) == 0 {
fmt.Println("No entries yet. Run `knox watch` to start ingesting.")
if page > 1 {
fmt.Printf("No more entries on page %d (page %d of %d).\n", page, page, (len(all)+limit-1)/limit)
} else {
fmt.Println("No entries yet. Run `knox watch` to start ingesting.")
}
return nil
}
fmt.Printf("Recent %d entries (page %d, %d per page):\n\n", len(entries), page, limit)