Files
hexis/internal/execution/audit_test.go
T
kami d4285607af Persist execution audit fields and enforce the in-flight guard in SQL
Findings 4 and 7 of REVIEW-2026-07-30.md, as a single user_version step
(11 -> 12) because they alter the same table.

confirmation_id was written by the INSERT and absent from all three executions
SELECTs, so it always read back empty: "which confirmation authorized this
destructive act" was unanswerable from the API. causation_id was set on the
struct by the engine with no column to land in, and capability_version was
missing entirely despite spec §4.1 and the precedent on confirmations. All
three are now persisted and selected back.

The one-in-flight-per-(capability, target) rule was check-then-insert with no
constraint between, so two concurrent requests could both proceed. It is now a
partial UNIQUE index; the engine's pre-insert query is demoted to an advisory
fast path and a constraint violation maps to ErrExecutionInFlight (409), kept
distinct from the idempotency-key replay case.

Note the migration also rewrites pre-existing duplicate in-flight rows to
status='unknown', keeping the lowest id per pair — creating the index would
otherwise fail outright on any database holding stale started rows, which the
old non-unique index permitted indefinitely. That is a write to existing audit
rows, not just DDL. Rows predating this migration get capability_version=0,
which is indistinguishable from a genuine 0; backfill is not possible.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Uea55zaiWuEByEDC4UBSdd
2026-07-30 23:39:26 +04:00

154 lines
4.7 KiB
Go

package execution_test
import (
"errors"
"sync"
"testing"
"time"
"github.com/kami/hexis/internal/domain"
)
// The audit trail is only useful if it survives a round trip through storage:
// "which confirmation authorized this destructive act" must be answerable from
// a plain read of the execution, not just from the in-memory struct the engine
// happened to return.
func TestExecute_AuditFieldsRoundTripThroughStorage(t *testing.T) {
prov := &fakeProvider{name: "fake"}
eng, store := newTestEngine(t, prov)
cap := mustCreateCapability(t, store, func(c *domain.Capability) {
c.RequiresConfirmation = true
c.Risk = "destructive"
})
args := map[string]any{"foo": "bar"}
conf, err := eng.CreateConfirmation(cap.ID, "ent_x", "kami", args)
if err != nil {
t.Fatalf("create confirmation: %v", err)
}
res, err := eng.Execute(&domain.ExecuteRequest{
CapabilityID: cap.ID,
TargetEntityID: "ent_x",
Arguments: args,
ConfirmationID: conf.ID,
CorrelationID: "corr_1",
CausationID: "caus_1",
IdempotencyKey: "idem_1",
})
if err != nil {
t.Fatalf("execute: %v", err)
}
// This is what GET /api/v1/executions/{id} serves.
got, err := store.GetExecution(res.Execution.ID)
if err != nil {
t.Fatalf("get execution: %v", err)
}
if got.ConfirmationID != conf.ID {
t.Errorf("confirmation_id: got %q, want %q", got.ConfirmationID, conf.ID)
}
if got.CausationID != "caus_1" {
t.Errorf("causation_id: got %q, want %q", got.CausationID, "caus_1")
}
if got.CorrelationID != "corr_1" {
t.Errorf("correlation_id: got %q, want %q", got.CorrelationID, "corr_1")
}
if got.CapabilityVersion != cap.Version {
t.Errorf("capability_version: got %d, want %d", got.CapabilityVersion, cap.Version)
}
// The idempotency-key read path serves the same record and must agree.
byKey, err := store.GetExecutionByIdempotencyKey("idem_1")
if err != nil {
t.Fatalf("get by idempotency key: %v", err)
}
if byKey.ConfirmationID != conf.ID || byKey.CausationID != "caus_1" || byKey.CapabilityVersion != cap.Version {
t.Errorf("idempotency read path lost audit fields: %+v", byKey)
}
}
// The in-flight guard used to be check-then-insert. Two concurrent identical
// executes must produce exactly one execution row, with the loser reported as
// already-in-flight rather than a 500.
func TestExecute_ConcurrentDuplicateProducesExactlyOneRow(t *testing.T) {
prov := &fakeProvider{name: "fake", delay: 150 * time.Millisecond}
eng, store := newTestEngine(t, prov)
cap := mustCreateCapability(t, store, nil)
const n = 4
var wg sync.WaitGroup
errs := make([]error, n)
start := make(chan struct{})
for i := 0; i < n; i++ {
wg.Add(1)
go func(i int) {
defer wg.Done()
<-start
_, errs[i] = eng.Execute(&domain.ExecuteRequest{
CapabilityID: cap.ID,
TargetEntityID: "ent_x",
})
}(i)
}
close(start)
wg.Wait()
succeeded := 0
for i, err := range errs {
switch {
case err == nil:
succeeded++
case errors.Is(err, domain.ErrExecutionInFlight):
default:
t.Errorf("goroutine %d: unexpected error %v (want nil or ErrExecutionInFlight)", i, err)
}
}
if succeeded != 1 {
t.Errorf("expected exactly 1 successful execute, got %d", succeeded)
}
var rows int
if err := store.DB().QueryRow(
`SELECT COUNT(*) FROM executions WHERE capability_id = ? AND target_entity_id = ?`,
cap.ID, "ent_x",
).Scan(&rows); err != nil {
t.Fatalf("count executions: %v", err)
}
if rows != 1 {
t.Errorf("expected exactly 1 execution row, got %d", rows)
}
}
// Directly exercise the partial unique index, bypassing the engine's advisory
// pre-check, so the constraint itself is what is under test.
func TestStore_SecondInFlightRowRejectedByIndex(t *testing.T) {
_, store := newTestEngine(t, &fakeProvider{name: "fake"})
now := time.Now().UTC()
mk := func(status domain.ExecutionStatus) *domain.Execution {
return &domain.Execution{
ID: domain.NewExecutionID(),
CapabilityID: "cap_1",
TargetEntityID: "ent_x",
Status: status,
CreatedAt: now,
UpdatedAt: now,
}
}
if err := store.CreateExecution(mk(domain.ExecutionStarted)); err != nil {
t.Fatalf("first insert: %v", err)
}
if err := store.CreateExecution(mk(domain.ExecutionStarted)); !errors.Is(err, domain.ErrExecutionInFlight) {
t.Fatalf("second in-flight insert: got %v, want ErrExecutionInFlight", err)
}
// The index is partial: terminal rows for the same pair stay insertable.
if err := store.CreateExecution(mk(domain.ExecutionSucceeded)); err != nil {
t.Fatalf("terminal-status insert should be allowed: %v", err)
}
if err := store.CreateExecution(mk(domain.ExecutionSucceeded)); err != nil {
t.Fatalf("second terminal-status insert should be allowed: %v", err)
}
}