devrites 3.2.2 → 3.2.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +8 -0
- package/README.md +1 -1
- package/docs/architecture.md +6 -2
- package/docs/command-map.md +3 -3
- package/docs/engine/agent-contract.md +3 -2
- package/docs/harness-compliance.md +2 -2
- package/docs/orchestration.md +15 -12
- package/docs/porting-to-a-new-harness.md +2 -1
- package/docs/skills.md +7 -2
- package/engine/hooks.go +39 -4
- package/engine/hooks_agent_dispatch.go +1193 -0
- package/engine/hooks_workspace.go +78 -10
- package/engine/internal/forge/forge.go +0 -2
- package/engine/internal/harness/compliance.go +1 -1
- package/engine/internal/harness/harness.go +3 -0
- package/engine/internal/lib/doubtcoverage.go +2 -12
- package/engine/internal/lib/reconcile.go +4 -29
- package/engine/internal/lib/reconcile_test.go +14 -10
- package/engine/main.go +4 -1
- package/engine/tests/hook_test.go +874 -1
- package/engine/tests/lib_parity_test.go +5 -17
- package/engine/tests/parity_reconcile_test.go +1 -1
- package/engine/tests/parity_test.go +2 -6
- package/pack/.claude/skills/devrites-api-interface/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-audit/SKILL.md +5 -6
- package/pack/.claude/skills/devrites-browser-proof/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-debug-recovery/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-doubt/SKILL.md +2 -1
- package/pack/.claude/skills/devrites-frontend-craft/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-interview/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-lib/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-lib/reference/model-tiers.md +4 -3
- package/pack/.claude/skills/devrites-lib/reference/parallel-dispatch.md +4 -5
- package/pack/.claude/skills/devrites-lib/reference/standards/agents.md +14 -8
- package/pack/.claude/skills/devrites-prose-craft/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-refresh-indexes/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-source-driven/SKILL.md +1 -0
- package/pack/.claude/skills/devrites-ux-shape/SKILL.md +1 -0
- package/pack/.claude/skills/rite/SKILL.md +1 -0
- package/pack/.claude/skills/rite-adopt/SKILL.md +1 -0
- package/pack/.claude/skills/rite-autocomplete/SKILL.md +1 -0
- package/pack/.claude/skills/rite-build/SKILL.md +2 -1
- package/pack/.claude/skills/rite-build/reference/phase-contract.md +3 -4
- package/pack/.claude/skills/rite-build/reference/wright-dispatch.md +11 -12
- package/pack/.claude/skills/rite-clarify/SKILL.md +1 -0
- package/pack/.claude/skills/rite-converge/SKILL.md +1 -0
- package/pack/.claude/skills/rite-customize/SKILL.md +1 -0
- package/pack/.claude/skills/rite-define/SKILL.md +1 -0
- package/pack/.claude/skills/rite-doctor/SKILL.md +1 -0
- package/pack/.claude/skills/rite-dogfood/SKILL.md +1 -0
- package/pack/.claude/skills/rite-explain/SKILL.md +4 -2
- package/pack/.claude/skills/rite-frame/SKILL.md +1 -0
- package/pack/.claude/skills/rite-handoff/SKILL.md +1 -0
- package/pack/.claude/skills/rite-learn/SKILL.md +1 -0
- package/pack/.claude/skills/rite-plan/SKILL.md +1 -0
- package/pack/.claude/skills/rite-polish/SKILL.md +1 -0
- package/pack/.claude/skills/rite-pov/SKILL.md +1 -0
- package/pack/.claude/skills/rite-pr-feedback/SKILL.md +1 -0
- package/pack/.claude/skills/rite-pressure-test/SKILL.md +1 -0
- package/pack/.claude/skills/rite-prototype/SKILL.md +1 -0
- package/pack/.claude/skills/rite-prove/SKILL.md +1 -0
- package/pack/.claude/skills/rite-quick/SKILL.md +1 -0
- package/pack/.claude/skills/rite-resolve/SKILL.md +1 -0
- package/pack/.claude/skills/rite-review/SKILL.md +1 -0
- package/pack/.claude/skills/rite-seal/SKILL.md +1 -0
- package/pack/.claude/skills/rite-seal/reference/phase-contract.md +2 -3
- package/pack/.claude/skills/rite-seal/reference/risk-and-rollback.md +10 -12
- package/pack/.claude/skills/rite-ship/SKILL.md +1 -0
- package/pack/.claude/skills/rite-spec/SKILL.md +1 -0
- package/pack/.claude/skills/rite-status/SKILL.md +1 -0
- package/pack/.claude/skills/rite-temper/SKILL.md +3 -2
- package/pack/.claude/skills/rite-upgrade/SKILL.md +3 -1
- package/pack/.claude/skills/rite-vet/SKILL.md +1 -0
- package/pack/.claude/skills/rite-zoom-out/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-api-interface/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-audit/SKILL.md +5 -6
- package/pack/generated/claude/skills/devrites-browser-proof/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-debug-recovery/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-doubt/SKILL.md +2 -1
- package/pack/generated/claude/skills/devrites-frontend-craft/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-interview/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-lib/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-lib/reference/model-tiers.md +4 -3
- package/pack/generated/claude/skills/devrites-lib/reference/parallel-dispatch.md +4 -5
- package/pack/generated/claude/skills/devrites-lib/reference/standards/agents.md +14 -8
- package/pack/generated/claude/skills/devrites-prose-craft/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-refresh-indexes/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-source-driven/SKILL.md +1 -0
- package/pack/generated/claude/skills/devrites-ux-shape/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-adopt/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-autocomplete/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-build/SKILL.md +2 -1
- package/pack/generated/claude/skills/rite-build/reference/phase-contract.md +3 -4
- package/pack/generated/claude/skills/rite-build/reference/wright-dispatch.md +11 -12
- package/pack/generated/claude/skills/rite-clarify/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-converge/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-customize/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-define/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-doctor/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-dogfood/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-explain/SKILL.md +4 -2
- package/pack/generated/claude/skills/rite-frame/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-handoff/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-learn/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-plan/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-polish/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-pov/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-pr-feedback/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-pressure-test/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-prototype/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-prove/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-quick/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-resolve/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-review/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-seal/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-seal/reference/phase-contract.md +2 -3
- package/pack/generated/claude/skills/rite-seal/reference/risk-and-rollback.md +10 -12
- package/pack/generated/claude/skills/rite-ship/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-spec/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-status/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-temper/SKILL.md +3 -2
- package/pack/generated/claude/skills/rite-upgrade/SKILL.md +3 -1
- package/pack/generated/claude/skills/rite-vet/SKILL.md +1 -0
- package/pack/generated/claude/skills/rite-zoom-out/SKILL.md +1 -0
- package/pack/generated/codex/AGENTS.md +2 -4
- package/pack/generated/codex/hooks.json +24 -5
- package/pack/generated/codex/skills/devrites-api-interface/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-audit/SKILL.md +10 -10
- package/pack/generated/codex/skills/devrites-browser-proof/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-debug-recovery/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-doubt/SKILL.md +7 -5
- package/pack/generated/codex/skills/devrites-frontend-craft/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-interview/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-lib/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-lib/reference/model-tiers.md +4 -3
- package/pack/generated/codex/skills/devrites-lib/reference/parallel-dispatch.md +4 -5
- package/pack/generated/codex/skills/devrites-lib/reference/standards/agents.md +14 -8
- package/pack/generated/codex/skills/devrites-prose-craft/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-refresh-indexes/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-source-driven/SKILL.md +6 -4
- package/pack/generated/codex/skills/devrites-ux-shape/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-adopt/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-autocomplete/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-build/SKILL.md +7 -5
- package/pack/generated/codex/skills/rite-build/reference/phase-contract.md +3 -4
- package/pack/generated/codex/skills/rite-build/reference/wright-dispatch.md +11 -12
- package/pack/generated/codex/skills/rite-clarify/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-converge/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-customize/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-define/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-doctor/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-dogfood/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-explain/SKILL.md +9 -6
- package/pack/generated/codex/skills/rite-frame/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-handoff/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-learn/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-plan/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-polish/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-pov/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-pr-feedback/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-pressure-test/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-prototype/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-prove/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-quick/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-resolve/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-review/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-seal/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-seal/reference/phase-contract.md +2 -3
- package/pack/generated/codex/skills/rite-seal/reference/risk-and-rollback.md +10 -12
- package/pack/generated/codex/skills/rite-ship/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-spec/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-status/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-temper/SKILL.md +8 -6
- package/pack/generated/codex/skills/rite-upgrade/SKILL.md +8 -5
- package/pack/generated/codex/skills/rite-vet/SKILL.md +6 -4
- package/pack/generated/codex/skills/rite-zoom-out/SKILL.md +6 -4
- package/package.json +1 -1
- package/scripts/codex-generate.sh +32 -18
- package/scripts/live-hosts/agent-result.schema.json +2 -1
- package/scripts/live-hosts/claude.sh +2 -2
- package/scripts/live-hosts/codex.sh +4 -0
- package/scripts/live-hosts/fake-host.py +1 -8
- package/scripts/live-hosts/host-transport.py +1 -2
- package/scripts/run-agent-contract-evals.py +3 -3
- package/scripts/run-evals.sh +3 -34
- package/scripts/validate-frontmatter.py +34 -1
- package/engine/testdata/golden/TestParityA1Guard/inline-fallback-silent.golden +0 -1
- package/engine/testdata/golden/TestParityDoubtCoverage/arg=inline.golden +0 -3
- package/engine/testdata/golden/TestParityDoubtCoverage/arg=missing-inline.golden +0 -3
- package/scripts/eval-runner.py +0 -275
- package/scripts/eval-scorer.mjs +0 -46
- package/scripts/render-eval-summary.py +0 -48
- /package/engine/testdata/golden/TestParityReconcile/{inline-fallback → close-window}/snapshot.golden +0 -0
- /package/engine/testdata/golden/TestParityReconcile/{inline-fallback.golden → close-window.golden} +0 -0
|
@@ -11,6 +11,7 @@ import (
|
|
|
11
11
|
"strconv"
|
|
12
12
|
"strings"
|
|
13
13
|
"testing"
|
|
14
|
+
"time"
|
|
14
15
|
)
|
|
15
16
|
|
|
16
17
|
// writeActive points the workspace's ACTIVE pointer at slug.
|
|
@@ -475,7 +476,7 @@ func TestHookWrightScopeBindsCodexWorkerDuringReconcileWindow(t *testing.T) {
|
|
|
475
476
|
if err := os.WriteFile(filepath.Join(workspace, ".reconcile-base"), []byte("snapshot\n"), 0o600); err != nil {
|
|
476
477
|
t.Fatal(err)
|
|
477
478
|
}
|
|
478
|
-
env := []string{"DEVRITES_CODEX_GENERIC_AGENT_COMPAT=1"}
|
|
479
|
+
env := []string{"DEVRITES_CODEX_GENERIC_AGENT_COMPAT=1", "DEVRITES_HOOK_PROFILE=strict"}
|
|
479
480
|
|
|
480
481
|
for _, test := range []struct {
|
|
481
482
|
name string
|
|
@@ -512,6 +513,137 @@ func TestHookWrightScopeBindsCodexWorkerDuringReconcileWindow(t *testing.T) {
|
|
|
512
513
|
}
|
|
513
514
|
}
|
|
514
515
|
|
|
516
|
+
func TestHookWrightScopeAcceptsCodexV2NamedRole(t *testing.T) {
|
|
517
|
+
root := newWorkspace(t)
|
|
518
|
+
writeActive(t, root, "auth-tokens")
|
|
519
|
+
workspace := filepath.Join(root, "features", "auth-tokens")
|
|
520
|
+
if err := os.WriteFile(filepath.Join(workspace, ".wright-allowlist"), []byte("src/app.go\n"), 0o600); err != nil {
|
|
521
|
+
t.Fatal(err)
|
|
522
|
+
}
|
|
523
|
+
for _, name := range []string{".reconcile-base", ".reconcile-allowlist", ".reconcile-devrites"} {
|
|
524
|
+
if err := os.WriteFile(filepath.Join(workspace, name), []byte(name+"\n"), 0o600); err != nil {
|
|
525
|
+
t.Fatal(err)
|
|
526
|
+
}
|
|
527
|
+
}
|
|
528
|
+
|
|
529
|
+
env := []string{"DEVRITES_HOOK_PROFILE=strict"}
|
|
530
|
+
for _, test := range []struct {
|
|
531
|
+
path string
|
|
532
|
+
deny bool
|
|
533
|
+
}{
|
|
534
|
+
{"src/app.go", false},
|
|
535
|
+
{"src/other.go", true},
|
|
536
|
+
} {
|
|
537
|
+
payload := hookPayload(t, map[string]any{
|
|
538
|
+
"hook_event_name": "PreToolUse",
|
|
539
|
+
"agent_type": "devrites-slice-wright",
|
|
540
|
+
"tool_name": "Edit",
|
|
541
|
+
"tool_input": map[string]any{"file_path": test.path},
|
|
542
|
+
})
|
|
543
|
+
out, stderr, code := runDevritesIO(t, root, payload, env,
|
|
544
|
+
"hook", "wright-scope", "--harness=codex")
|
|
545
|
+
if code != 0 {
|
|
546
|
+
t.Fatalf("%s exit=%d stderr=%s", test.path, code, stderr)
|
|
547
|
+
}
|
|
548
|
+
if test.deny {
|
|
549
|
+
if decision, _ := parsePermissionDecision(t, out); decision != "deny" {
|
|
550
|
+
t.Fatalf("%s should be denied: %s", test.path, out)
|
|
551
|
+
}
|
|
552
|
+
} else if strings.TrimSpace(out) != "" {
|
|
553
|
+
t.Fatalf("%s should be allowed: %s", test.path, out)
|
|
554
|
+
}
|
|
555
|
+
}
|
|
556
|
+
}
|
|
557
|
+
|
|
558
|
+
func TestHookWrightScopeRequiresSpawnedCodexWorkerForSourceWrites(t *testing.T) {
|
|
559
|
+
root := newWorkspace(t)
|
|
560
|
+
writeActive(t, root, "auth-tokens")
|
|
561
|
+
workspace := filepath.Join(root, "features", "auth-tokens")
|
|
562
|
+
if err := os.WriteFile(filepath.Join(workspace, ".reconcile-base"), []byte("snapshot\n"), 0o600); err != nil {
|
|
563
|
+
t.Fatal(err)
|
|
564
|
+
}
|
|
565
|
+
env := []string{"DEVRITES_CODEX_GENERIC_AGENT_COMPAT=1"}
|
|
566
|
+
|
|
567
|
+
runGuard := func(input string, deny bool) {
|
|
568
|
+
t.Helper()
|
|
569
|
+
out, errOut, code := runDevritesIO(t, root, input, env,
|
|
570
|
+
"hook", "wright-scope", "--harness=codex")
|
|
571
|
+
if code != 0 {
|
|
572
|
+
t.Fatalf("exit=%d out=%q stderr=%q", code, out, errOut)
|
|
573
|
+
}
|
|
574
|
+
if deny {
|
|
575
|
+
if decision, _ := parsePermissionDecision(t, out); decision != "deny" {
|
|
576
|
+
t.Fatalf("expected deny, got %q", out)
|
|
577
|
+
}
|
|
578
|
+
} else if strings.TrimSpace(out) != "" {
|
|
579
|
+
t.Fatalf("expected allow, got %q", out)
|
|
580
|
+
}
|
|
581
|
+
}
|
|
582
|
+
|
|
583
|
+
runGuard(`{"tool_name":"spawn_agent","tool_input":{"agent_type":"worker"}}`, false)
|
|
584
|
+
runGuard(`{"tool_name":"spawn_agent","tool_input":{"agent_type":"explorer"}}`, false)
|
|
585
|
+
runGuard(`{"tool_name":"Write","tool_input":{"file_path":".devrites/features/auth-tokens/state.md"}}`, false)
|
|
586
|
+
runGuard(`{"tool_name":"Write","tool_input":{"file_path":"src/app.go"}}`, true)
|
|
587
|
+
runGuard(`{"tool_name":"Bash","tool_input":{"command":"printf x > src/app.go"}}`, true)
|
|
588
|
+
runGuard(`{"tool_name":"js","tool_input":{"code":"writeFile()"}}`, true)
|
|
589
|
+
out, errOut, code := runDevritesIO(t, root,
|
|
590
|
+
`{"tool_name":"Bash","tool_input":{"command":"printf x > src/app.go"}}`,
|
|
591
|
+
nil, "hook", "wright-scope", "--harness=claude")
|
|
592
|
+
if code != 0 {
|
|
593
|
+
t.Fatalf("Claude root guard: exit=%d stderr=%q", code, errOut)
|
|
594
|
+
}
|
|
595
|
+
if decision, _ := parsePermissionDecision(t, out); decision != "deny" {
|
|
596
|
+
t.Fatalf("Claude root source write was not denied: %q", out)
|
|
597
|
+
}
|
|
598
|
+
|
|
599
|
+
// A legacy marker is inert: it cannot authorize root source writes.
|
|
600
|
+
if err := os.WriteFile(filepath.Join(workspace, ".reconcile-inline"), nil, 0o600); err != nil {
|
|
601
|
+
t.Fatal(err)
|
|
602
|
+
}
|
|
603
|
+
runGuard(`{"tool_name":"Bash","tool_input":{"command":"printf x > src/app.go"}}`, true)
|
|
604
|
+
runGuard(`{"tool_name":"js","tool_input":{"code":"writeFile()"}}`, true)
|
|
605
|
+
|
|
606
|
+
projectDir := filepath.Dir(root)
|
|
607
|
+
if err := os.MkdirAll(filepath.Join(projectDir, "src"), 0o755); err != nil {
|
|
608
|
+
t.Fatal(err)
|
|
609
|
+
}
|
|
610
|
+
source := filepath.Join(projectDir, "src", "app.go")
|
|
611
|
+
if err := os.WriteFile(source, []byte("package src\n"), 0o600); err != nil {
|
|
612
|
+
t.Fatal(err)
|
|
613
|
+
}
|
|
614
|
+
alias := filepath.Join(workspace, "source-alias")
|
|
615
|
+
if err := os.Symlink(source, alias); err != nil {
|
|
616
|
+
t.Skipf("symlinks unavailable: %v", err)
|
|
617
|
+
}
|
|
618
|
+
runGuard(`{"tool_name":"Write","tool_input":{"file_path":".devrites/features/auth-tokens/source-alias"}}`, true)
|
|
619
|
+
}
|
|
620
|
+
|
|
621
|
+
func TestHookA1GuardIgnoresLegacyInlineMarker(t *testing.T) {
|
|
622
|
+
root := newWorkspace(t)
|
|
623
|
+
writeActive(t, root, "auth-tokens")
|
|
624
|
+
workspace := filepath.Join(root, "features", "auth-tokens")
|
|
625
|
+
for _, name := range []string{".reconcile-base", ".reconcile-inline"} {
|
|
626
|
+
if err := os.WriteFile(filepath.Join(workspace, name), nil, 0o600); err != nil {
|
|
627
|
+
t.Fatal(err)
|
|
628
|
+
}
|
|
629
|
+
}
|
|
630
|
+
input := `{"tool_name":"Edit","tool_input":{"file_path":"src/app.go"}}`
|
|
631
|
+
env := []string{"DEVRITES_A1_HOOK=enforce"}
|
|
632
|
+
|
|
633
|
+
for _, host := range []string{"codex", "claude"} {
|
|
634
|
+
t.Run(host, func(t *testing.T) {
|
|
635
|
+
out, errOut, code := runDevritesIO(t, root, input, env,
|
|
636
|
+
"hook", "a1-guard", "--harness="+host)
|
|
637
|
+
if code != 0 {
|
|
638
|
+
t.Fatalf("exit=%d stderr=%q", code, errOut)
|
|
639
|
+
}
|
|
640
|
+
if decision, _ := parsePermissionDecision(t, out); decision != "deny" {
|
|
641
|
+
t.Fatalf("legacy inline marker authorized root source write: %q", out)
|
|
642
|
+
}
|
|
643
|
+
})
|
|
644
|
+
}
|
|
645
|
+
}
|
|
646
|
+
|
|
515
647
|
func TestHookWrightScopeFailsClosedWhenWorkspaceResolutionFails(t *testing.T) {
|
|
516
648
|
root := newWorkspace(t)
|
|
517
649
|
if err := os.Remove(filepath.Join(root, "ACTIVE")); err != nil && !os.IsNotExist(err) {
|
|
@@ -733,3 +865,744 @@ func TestHookSubagentOrientSilentForNonDevritesAgent(t *testing.T) {
|
|
|
733
865
|
}
|
|
734
866
|
}
|
|
735
867
|
}
|
|
868
|
+
|
|
869
|
+
func hookPayload(t *testing.T, value map[string]any) string {
|
|
870
|
+
t.Helper()
|
|
871
|
+
raw, err := json.Marshal(value)
|
|
872
|
+
if err != nil {
|
|
873
|
+
t.Fatal(err)
|
|
874
|
+
}
|
|
875
|
+
return string(raw)
|
|
876
|
+
}
|
|
877
|
+
|
|
878
|
+
func writeCodexAgentContract(t *testing.T, root, role string) {
|
|
879
|
+
t.Helper()
|
|
880
|
+
dir := filepath.Join(filepath.Dir(root), ".codex", "agents")
|
|
881
|
+
if err := os.MkdirAll(dir, 0o755); err != nil {
|
|
882
|
+
t.Fatal(err)
|
|
883
|
+
}
|
|
884
|
+
contract := "name = \"" + role + "\"\ndeveloper_instructions = '''\nROLE-CONTRACT:" + role + "\n'''\n"
|
|
885
|
+
if err := os.WriteFile(filepath.Join(dir, role+".toml"), []byte(contract), 0o600); err != nil {
|
|
886
|
+
t.Fatal(err)
|
|
887
|
+
}
|
|
888
|
+
}
|
|
889
|
+
|
|
890
|
+
func writeCodexSkillContract(t *testing.T, root, skill, roles string) {
|
|
891
|
+
t.Helper()
|
|
892
|
+
dir := filepath.Join(filepath.Dir(root), ".agents", "skills", skill)
|
|
893
|
+
if err := os.MkdirAll(dir, 0o755); err != nil {
|
|
894
|
+
t.Fatal(err)
|
|
895
|
+
}
|
|
896
|
+
contract := "---\nname: " + skill + "\ndescription: Test skill.\nrequired-agent-roles: " + roles + "\n---\n"
|
|
897
|
+
if err := os.WriteFile(filepath.Join(dir, "SKILL.md"), []byte(contract), 0o600); err != nil {
|
|
898
|
+
t.Fatal(err)
|
|
899
|
+
}
|
|
900
|
+
}
|
|
901
|
+
|
|
902
|
+
func writeCodexV2Rollouts(
|
|
903
|
+
t *testing.T,
|
|
904
|
+
codeHome, projectRoot, sessionID, turnID, role, taskName string,
|
|
905
|
+
includeWait, includeInstructions, includeResult bool,
|
|
906
|
+
) {
|
|
907
|
+
t.Helper()
|
|
908
|
+
dir := filepath.Join(codeHome, "sessions", "2026", "07", "25")
|
|
909
|
+
if err := os.MkdirAll(dir, 0o700); err != nil {
|
|
910
|
+
t.Fatal(err)
|
|
911
|
+
}
|
|
912
|
+
at := time.Now().UTC()
|
|
913
|
+
parent := []map[string]any{
|
|
914
|
+
{
|
|
915
|
+
"timestamp": at.Format(time.RFC3339Nano),
|
|
916
|
+
"type": "session_meta",
|
|
917
|
+
"payload": map[string]any{
|
|
918
|
+
"session_id": sessionID,
|
|
919
|
+
"id": sessionID,
|
|
920
|
+
"cwd": projectRoot,
|
|
921
|
+
},
|
|
922
|
+
},
|
|
923
|
+
{
|
|
924
|
+
"timestamp": at.Add(time.Millisecond).Format(time.RFC3339Nano),
|
|
925
|
+
"type": "response_item",
|
|
926
|
+
"payload": map[string]any{
|
|
927
|
+
"type": "function_call",
|
|
928
|
+
"name": "spawn_agent",
|
|
929
|
+
"arguments": `{"agent_type":"` + role + `","task_name":"` + taskName + `","fork_turns":"none","message":"encrypted"}`,
|
|
930
|
+
"internal_chat_message_metadata_passthrough": map[string]any{
|
|
931
|
+
"turn_id": turnID,
|
|
932
|
+
},
|
|
933
|
+
},
|
|
934
|
+
},
|
|
935
|
+
}
|
|
936
|
+
if includeWait {
|
|
937
|
+
parent = append(parent, map[string]any{
|
|
938
|
+
"timestamp": at.Add(2 * time.Millisecond).Format(time.RFC3339Nano),
|
|
939
|
+
"type": "response_item",
|
|
940
|
+
"payload": map[string]any{
|
|
941
|
+
"type": "function_call",
|
|
942
|
+
"name": "wait_agent",
|
|
943
|
+
"arguments": `{"timeout_ms":3600000}`,
|
|
944
|
+
"internal_chat_message_metadata_passthrough": map[string]any{
|
|
945
|
+
"turn_id": turnID,
|
|
946
|
+
},
|
|
947
|
+
},
|
|
948
|
+
})
|
|
949
|
+
}
|
|
950
|
+
if includeResult {
|
|
951
|
+
parent = append(parent, map[string]any{
|
|
952
|
+
"timestamp": at.Add(3 * time.Millisecond).Format(time.RFC3339Nano),
|
|
953
|
+
"type": "response_item",
|
|
954
|
+
"payload": map[string]any{
|
|
955
|
+
"type": "agent_message",
|
|
956
|
+
"author": "/root/" + taskName,
|
|
957
|
+
"recipient": "/root",
|
|
958
|
+
"content": []map[string]any{{
|
|
959
|
+
"type": "input_text",
|
|
960
|
+
"text": "Message Type: FINAL_ANSWER\nPayload:\nagent-result/v1 complete",
|
|
961
|
+
}},
|
|
962
|
+
"internal_chat_message_metadata_passthrough": map[string]any{
|
|
963
|
+
"turn_id": turnID,
|
|
964
|
+
},
|
|
965
|
+
},
|
|
966
|
+
})
|
|
967
|
+
}
|
|
968
|
+
writeJSONLines(t, filepath.Join(dir, "rollout-parent-"+sessionID+".jsonl"), parent)
|
|
969
|
+
|
|
970
|
+
child := []map[string]any{
|
|
971
|
+
{
|
|
972
|
+
"timestamp": at.Add(time.Millisecond).Format(time.RFC3339Nano),
|
|
973
|
+
"type": "session_meta",
|
|
974
|
+
"payload": map[string]any{
|
|
975
|
+
"session_id": sessionID,
|
|
976
|
+
"id": "child-" + taskName,
|
|
977
|
+
"parent_thread_id": sessionID,
|
|
978
|
+
"cwd": projectRoot,
|
|
979
|
+
"agent_path": "/root/" + taskName,
|
|
980
|
+
"agent_role": role,
|
|
981
|
+
},
|
|
982
|
+
},
|
|
983
|
+
}
|
|
984
|
+
if includeInstructions {
|
|
985
|
+
child = append(child, map[string]any{
|
|
986
|
+
"timestamp": at.Add(2 * time.Millisecond).Format(time.RFC3339Nano),
|
|
987
|
+
"type": "response_item",
|
|
988
|
+
"payload": map[string]any{
|
|
989
|
+
"type": "message",
|
|
990
|
+
"role": "developer",
|
|
991
|
+
"content": []map[string]any{{
|
|
992
|
+
"type": "input_text",
|
|
993
|
+
"text": "ROLE-CONTRACT:" + role,
|
|
994
|
+
}},
|
|
995
|
+
},
|
|
996
|
+
})
|
|
997
|
+
}
|
|
998
|
+
if includeResult {
|
|
999
|
+
child = append(child, map[string]any{
|
|
1000
|
+
"timestamp": at.Add(3 * time.Millisecond).Format(time.RFC3339Nano),
|
|
1001
|
+
"type": "event_msg",
|
|
1002
|
+
"payload": map[string]any{
|
|
1003
|
+
"type": "task_complete",
|
|
1004
|
+
"last_agent_message": "agent-result/v1 complete",
|
|
1005
|
+
},
|
|
1006
|
+
})
|
|
1007
|
+
}
|
|
1008
|
+
writeJSONLines(t, filepath.Join(dir, "rollout-child-"+taskName+".jsonl"), child)
|
|
1009
|
+
}
|
|
1010
|
+
|
|
1011
|
+
func writeJSONLines(t *testing.T, path string, records []map[string]any) {
|
|
1012
|
+
t.Helper()
|
|
1013
|
+
var lines []string
|
|
1014
|
+
for _, record := range records {
|
|
1015
|
+
raw, err := json.Marshal(record)
|
|
1016
|
+
if err != nil {
|
|
1017
|
+
t.Fatal(err)
|
|
1018
|
+
}
|
|
1019
|
+
lines = append(lines, string(raw))
|
|
1020
|
+
}
|
|
1021
|
+
if err := os.WriteFile(path, []byte(strings.Join(lines, "\n")+"\n"), 0o600); err != nil {
|
|
1022
|
+
t.Fatal(err)
|
|
1023
|
+
}
|
|
1024
|
+
}
|
|
1025
|
+
|
|
1026
|
+
func TestCodexAgentDispatchArmsInstalledSkillRequirements(t *testing.T) {
|
|
1027
|
+
for _, tc := range []struct {
|
|
1028
|
+
name string
|
|
1029
|
+
invocation string
|
|
1030
|
+
roles string
|
|
1031
|
+
wantRoles []string
|
|
1032
|
+
}{
|
|
1033
|
+
{
|
|
1034
|
+
name: "single-role-dollar-invocation",
|
|
1035
|
+
invocation: "$rite-vet",
|
|
1036
|
+
roles: "devrites-plan-reviewer",
|
|
1037
|
+
wantRoles: []string{"devrites-plan-reviewer"},
|
|
1038
|
+
},
|
|
1039
|
+
{
|
|
1040
|
+
name: "multi-role-slash-invocation",
|
|
1041
|
+
invocation: "/rite-review",
|
|
1042
|
+
roles: "devrites-spec-reviewer, devrites-code-reviewer",
|
|
1043
|
+
wantRoles: []string{"devrites-code-reviewer", "devrites-spec-reviewer"},
|
|
1044
|
+
},
|
|
1045
|
+
{
|
|
1046
|
+
name: "declared-no-agent",
|
|
1047
|
+
invocation: "$rite-status",
|
|
1048
|
+
roles: "none",
|
|
1049
|
+
},
|
|
1050
|
+
} {
|
|
1051
|
+
t.Run(tc.name, func(t *testing.T) {
|
|
1052
|
+
root := newWorkspace(t)
|
|
1053
|
+
skill := strings.TrimLeft(tc.invocation, "$/")
|
|
1054
|
+
writeCodexSkillContract(t, root, skill, tc.roles)
|
|
1055
|
+
for _, role := range tc.wantRoles {
|
|
1056
|
+
writeCodexAgentContract(t, root, role)
|
|
1057
|
+
}
|
|
1058
|
+
sessionID, turnID := "session-"+tc.name, "turn-"+tc.name
|
|
1059
|
+
out, stderr, code := runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1060
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1061
|
+
"session_id": sessionID,
|
|
1062
|
+
"turn_id": turnID,
|
|
1063
|
+
"prompt": tc.invocation,
|
|
1064
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1065
|
+
if code != 0 {
|
|
1066
|
+
t.Fatalf("arm exit=%d stderr=%s", code, stderr)
|
|
1067
|
+
}
|
|
1068
|
+
for _, role := range tc.wantRoles {
|
|
1069
|
+
if !strings.Contains(out, role) {
|
|
1070
|
+
t.Fatalf("arm output omitted %s: %s", role, out)
|
|
1071
|
+
}
|
|
1072
|
+
}
|
|
1073
|
+
|
|
1074
|
+
stopOut, stopErr, stopCode := runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1075
|
+
"hook_event_name": "Stop",
|
|
1076
|
+
"session_id": sessionID,
|
|
1077
|
+
"turn_id": turnID,
|
|
1078
|
+
"stop_hook_active": false,
|
|
1079
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1080
|
+
if stopCode != 0 {
|
|
1081
|
+
t.Fatalf("stop exit=%d stderr=%s", stopCode, stopErr)
|
|
1082
|
+
}
|
|
1083
|
+
if len(tc.wantRoles) == 0 {
|
|
1084
|
+
if strings.TrimSpace(stopOut) != "" {
|
|
1085
|
+
t.Fatalf("no-agent skill was blocked: %s", stopOut)
|
|
1086
|
+
}
|
|
1087
|
+
return
|
|
1088
|
+
}
|
|
1089
|
+
for _, role := range tc.wantRoles {
|
|
1090
|
+
if strings.Contains(stopOut, role) {
|
|
1091
|
+
return
|
|
1092
|
+
}
|
|
1093
|
+
}
|
|
1094
|
+
t.Fatalf("required skill roles did not block Stop: %s", stopOut)
|
|
1095
|
+
})
|
|
1096
|
+
}
|
|
1097
|
+
}
|
|
1098
|
+
|
|
1099
|
+
func TestCodexAgentDispatchRejectsInvalidInstalledSkillRequirements(t *testing.T) {
|
|
1100
|
+
root := newWorkspace(t)
|
|
1101
|
+
writeCodexSkillContract(t, root, "rite-vet", "root-inline-reviewer")
|
|
1102
|
+
out, stderr, code := runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1103
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1104
|
+
"session_id": "session-invalid-skill",
|
|
1105
|
+
"turn_id": "turn-invalid-skill",
|
|
1106
|
+
"prompt": "$rite-vet",
|
|
1107
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1108
|
+
if code != 0 {
|
|
1109
|
+
t.Fatalf("invalid metadata exit=%d stderr=%s", code, stderr)
|
|
1110
|
+
}
|
|
1111
|
+
var decision struct {
|
|
1112
|
+
Decision string `json:"decision"`
|
|
1113
|
+
Reason string `json:"reason"`
|
|
1114
|
+
}
|
|
1115
|
+
if err := json.Unmarshal([]byte(strings.TrimSpace(out)), &decision); err != nil {
|
|
1116
|
+
t.Fatalf("invalid metadata response: %v\n%s", err, out)
|
|
1117
|
+
}
|
|
1118
|
+
if decision.Decision != "block" || !strings.Contains(decision.Reason, "invalid required agent role") {
|
|
1119
|
+
t.Fatalf("invalid skill metadata did not fail closed: %#v", decision)
|
|
1120
|
+
}
|
|
1121
|
+
stopOut, stopErr, stopCode := runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1122
|
+
"hook_event_name": "Stop",
|
|
1123
|
+
"session_id": "session-invalid-skill",
|
|
1124
|
+
"turn_id": "turn-invalid-skill",
|
|
1125
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1126
|
+
if stopCode != 0 || !strings.Contains(stopOut, "required-agent-roles contract is invalid") {
|
|
1127
|
+
t.Fatalf("invalid skill metadata did not remain fail closed at Stop: exit=%d out=%s stderr=%s", stopCode, stopOut, stopErr)
|
|
1128
|
+
}
|
|
1129
|
+
}
|
|
1130
|
+
|
|
1131
|
+
func TestCodexAgentDispatchDoesNotArmSkillNamesInsideURLs(t *testing.T) {
|
|
1132
|
+
root := newWorkspace(t)
|
|
1133
|
+
writeCodexSkillContract(t, root, "rite-vet", "devrites-plan-reviewer")
|
|
1134
|
+
sessionID, turnID := "session-skill-url", "turn-skill-url"
|
|
1135
|
+
runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1136
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1137
|
+
"session_id": sessionID,
|
|
1138
|
+
"turn_id": turnID,
|
|
1139
|
+
"prompt": "Read https://example.test/rite-vet before answering.",
|
|
1140
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1141
|
+
out, stderr, code := runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1142
|
+
"hook_event_name": "Stop",
|
|
1143
|
+
"session_id": sessionID,
|
|
1144
|
+
"turn_id": turnID,
|
|
1145
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1146
|
+
if code != 0 || strings.TrimSpace(out) != "" {
|
|
1147
|
+
t.Fatalf("URL path armed a skill: exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1148
|
+
}
|
|
1149
|
+
}
|
|
1150
|
+
|
|
1151
|
+
func TestCodexAgentDispatchDoesNotArmBareAgentContractReference(t *testing.T) {
|
|
1152
|
+
root := newWorkspace(t)
|
|
1153
|
+
writeCodexAgentContract(t, root, "devrites-security-auditor")
|
|
1154
|
+
sessionID, turnID := "session-agent-reference", "turn-agent-reference"
|
|
1155
|
+
runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1156
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1157
|
+
"session_id": sessionID,
|
|
1158
|
+
"turn_id": turnID,
|
|
1159
|
+
"prompt": "Explain .codex/agents/devrites-security-auditor.toml without running it.",
|
|
1160
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1161
|
+
out, stderr, code := runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1162
|
+
"hook_event_name": "Stop",
|
|
1163
|
+
"session_id": sessionID,
|
|
1164
|
+
"turn_id": turnID,
|
|
1165
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1166
|
+
if code != 0 || strings.TrimSpace(out) != "" {
|
|
1167
|
+
t.Fatalf("bare agent reference armed dispatch: exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1168
|
+
}
|
|
1169
|
+
}
|
|
1170
|
+
|
|
1171
|
+
func TestCodexAgentDispatchBlocksFalseWaitAndStop(t *testing.T) {
|
|
1172
|
+
root := newWorkspace(t)
|
|
1173
|
+
writeCodexSkillContract(t, root, "devrites-audit", "devrites-security-auditor")
|
|
1174
|
+
sessionID, turnID := "session-false-wait", "turn-false-wait"
|
|
1175
|
+
prompt := hookPayload(t, map[string]any{
|
|
1176
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1177
|
+
"session_id": sessionID,
|
|
1178
|
+
"turn_id": turnID,
|
|
1179
|
+
"prompt": "$devrites-audit Call spawn_agent and tell it to read .codex/agents/devrites-security-auditor.toml.",
|
|
1180
|
+
})
|
|
1181
|
+
if _, stderr, code := runDevritesIO(t, root, prompt, nil,
|
|
1182
|
+
"hook", "agent-dispatch", "--harness=codex"); code != 0 {
|
|
1183
|
+
t.Fatalf("arm exit=%d stderr=%s", code, stderr)
|
|
1184
|
+
}
|
|
1185
|
+
|
|
1186
|
+
wait := hookPayload(t, map[string]any{
|
|
1187
|
+
"hook_event_name": "PreToolUse",
|
|
1188
|
+
"session_id": sessionID,
|
|
1189
|
+
"turn_id": turnID,
|
|
1190
|
+
"tool_name": "wait",
|
|
1191
|
+
"tool_input": map[string]any{"receiver_thread_ids": []string{}},
|
|
1192
|
+
})
|
|
1193
|
+
out, stderr, code := runDevritesIO(t, root, wait, nil,
|
|
1194
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1195
|
+
if code != 0 {
|
|
1196
|
+
t.Fatalf("wait exit=%d stderr=%s", code, stderr)
|
|
1197
|
+
}
|
|
1198
|
+
if decision, reason := parsePermissionDecision(t, out); decision != "deny" ||
|
|
1199
|
+
!strings.Contains(reason, "spawn_agent") {
|
|
1200
|
+
t.Fatalf("false wait not denied: decision=%q reason=%q out=%s", decision, reason, out)
|
|
1201
|
+
}
|
|
1202
|
+
|
|
1203
|
+
stop := hookPayload(t, map[string]any{
|
|
1204
|
+
"hook_event_name": "Stop",
|
|
1205
|
+
"session_id": sessionID,
|
|
1206
|
+
"turn_id": turnID,
|
|
1207
|
+
"stop_hook_active": false,
|
|
1208
|
+
})
|
|
1209
|
+
out, stderr, code = runDevritesIO(t, root, stop, nil,
|
|
1210
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1211
|
+
if code != 0 {
|
|
1212
|
+
t.Fatalf("stop exit=%d stderr=%s", code, stderr)
|
|
1213
|
+
}
|
|
1214
|
+
var stopDecision struct {
|
|
1215
|
+
Decision string `json:"decision"`
|
|
1216
|
+
Reason string `json:"reason"`
|
|
1217
|
+
}
|
|
1218
|
+
if err := json.Unmarshal([]byte(strings.TrimSpace(out)), &stopDecision); err != nil {
|
|
1219
|
+
t.Fatalf("invalid Stop response: %v\n%s", err, out)
|
|
1220
|
+
}
|
|
1221
|
+
if stopDecision.Decision != "block" || !strings.Contains(stopDecision.Reason, "spawn_agent") {
|
|
1222
|
+
t.Fatalf("false completion not blocked: %#v", stopDecision)
|
|
1223
|
+
}
|
|
1224
|
+
|
|
1225
|
+
stop = hookPayload(t, map[string]any{
|
|
1226
|
+
"hook_event_name": "Stop",
|
|
1227
|
+
"session_id": sessionID,
|
|
1228
|
+
"turn_id": turnID,
|
|
1229
|
+
"stop_hook_active": true,
|
|
1230
|
+
})
|
|
1231
|
+
out, stderr, code = runDevritesIO(t, root, stop, nil,
|
|
1232
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1233
|
+
if code != 0 {
|
|
1234
|
+
t.Fatalf("second stop exit=%d stderr=%s", code, stderr)
|
|
1235
|
+
}
|
|
1236
|
+
var terminal struct {
|
|
1237
|
+
Continue bool `json:"continue"`
|
|
1238
|
+
StopReason string `json:"stopReason"`
|
|
1239
|
+
}
|
|
1240
|
+
if err := json.Unmarshal([]byte(strings.TrimSpace(out)), &terminal); err != nil {
|
|
1241
|
+
t.Fatalf("invalid terminal Stop response: %v\n%s", err, out)
|
|
1242
|
+
}
|
|
1243
|
+
if terminal.Continue || !strings.Contains(terminal.StopReason, "spawn_agent") {
|
|
1244
|
+
t.Fatalf("second false completion was not stopped: %#v", terminal)
|
|
1245
|
+
}
|
|
1246
|
+
}
|
|
1247
|
+
|
|
1248
|
+
func TestCodexAgentDispatchConfirmsGenericRoleAndResult(t *testing.T) {
|
|
1249
|
+
root := newWorkspace(t)
|
|
1250
|
+
sessionID, turnID := "session-complete", "turn-complete"
|
|
1251
|
+
role := "devrites-security-auditor"
|
|
1252
|
+
writeCodexAgentContract(t, root, role)
|
|
1253
|
+
writeCodexSkillContract(t, root, "devrites-audit", role)
|
|
1254
|
+
|
|
1255
|
+
prompt := hookPayload(t, map[string]any{
|
|
1256
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1257
|
+
"session_id": sessionID,
|
|
1258
|
+
"turn_id": turnID,
|
|
1259
|
+
"prompt": "$devrites-audit Call spawn_agent and tell it to read .codex/agents/" + role + ".toml.",
|
|
1260
|
+
})
|
|
1261
|
+
runDevritesIO(t, root, prompt, nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1262
|
+
|
|
1263
|
+
spawn := hookPayload(t, map[string]any{
|
|
1264
|
+
"hook_event_name": "PreToolUse",
|
|
1265
|
+
"session_id": sessionID,
|
|
1266
|
+
"turn_id": turnID,
|
|
1267
|
+
"tool_name": "spawn_agent",
|
|
1268
|
+
"tool_use_id": "spawn-1",
|
|
1269
|
+
"tool_input": map[string]any{
|
|
1270
|
+
"agent_type": "explorer",
|
|
1271
|
+
"fork_turns": "none",
|
|
1272
|
+
"message": "Read .codex/agents/" + role + ".toml and return the requested report.",
|
|
1273
|
+
},
|
|
1274
|
+
})
|
|
1275
|
+
out, stderr, code := runDevritesIO(t, root, spawn, nil,
|
|
1276
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1277
|
+
if code != 0 {
|
|
1278
|
+
t.Fatalf("spawn exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1279
|
+
}
|
|
1280
|
+
var rewrite struct {
|
|
1281
|
+
HookSpecificOutput struct {
|
|
1282
|
+
PermissionDecision string `json:"permissionDecision"`
|
|
1283
|
+
UpdatedInput struct {
|
|
1284
|
+
Message string `json:"message"`
|
|
1285
|
+
} `json:"updatedInput"`
|
|
1286
|
+
} `json:"hookSpecificOutput"`
|
|
1287
|
+
}
|
|
1288
|
+
if err := json.Unmarshal([]byte(strings.TrimSpace(out)), &rewrite); err != nil {
|
|
1289
|
+
t.Fatalf("invalid spawn rewrite: %v\n%s", err, out)
|
|
1290
|
+
}
|
|
1291
|
+
if rewrite.HookSpecificOutput.PermissionDecision != "allow" ||
|
|
1292
|
+
!strings.Contains(rewrite.HookSpecificOutput.UpdatedInput.Message, "ROLE-CONTRACT:"+role) ||
|
|
1293
|
+
!strings.Contains(rewrite.HookSpecificOutput.UpdatedInput.Message, "return the requested report") {
|
|
1294
|
+
t.Fatalf("generic spawn did not receive developer_instructions: %#v", rewrite)
|
|
1295
|
+
}
|
|
1296
|
+
|
|
1297
|
+
start := hookPayload(t, map[string]any{
|
|
1298
|
+
"hook_event_name": "SubagentStart",
|
|
1299
|
+
"session_id": sessionID,
|
|
1300
|
+
"turn_id": "child-turn",
|
|
1301
|
+
"agent_id": "agent-1",
|
|
1302
|
+
"agent_type": "explorer",
|
|
1303
|
+
})
|
|
1304
|
+
out, stderr, code = runDevritesIO(t, root, start, nil,
|
|
1305
|
+
"hook", "subagent-orient", "--harness=codex")
|
|
1306
|
+
if code != 0 {
|
|
1307
|
+
t.Fatalf("start exit=%d stderr=%s", code, stderr)
|
|
1308
|
+
}
|
|
1309
|
+
var startEnvelope struct {
|
|
1310
|
+
HookSpecificOutput struct {
|
|
1311
|
+
AdditionalContext string `json:"additionalContext"`
|
|
1312
|
+
} `json:"hookSpecificOutput"`
|
|
1313
|
+
}
|
|
1314
|
+
if err := json.Unmarshal([]byte(strings.TrimSpace(out)), &startEnvelope); err != nil {
|
|
1315
|
+
t.Fatalf("invalid start response: %v\n%s", err, out)
|
|
1316
|
+
}
|
|
1317
|
+
if !strings.Contains(startEnvelope.HookSpecificOutput.AdditionalContext,
|
|
1318
|
+
".codex/agents/"+role+".toml") {
|
|
1319
|
+
t.Fatalf("generic child did not receive role contract: %s", out)
|
|
1320
|
+
}
|
|
1321
|
+
|
|
1322
|
+
stopAgent := hookPayload(t, map[string]any{
|
|
1323
|
+
"hook_event_name": "SubagentStop",
|
|
1324
|
+
"session_id": sessionID,
|
|
1325
|
+
"turn_id": "child-turn",
|
|
1326
|
+
"agent_id": "agent-1",
|
|
1327
|
+
"agent_type": "explorer",
|
|
1328
|
+
"last_assistant_message": "agent-result/v1 complete",
|
|
1329
|
+
})
|
|
1330
|
+
if out, stderr, code := runDevritesIO(t, root, stopAgent, nil,
|
|
1331
|
+
"hook", "agent-dispatch", "--harness=codex"); code != 0 || strings.TrimSpace(out) != "" {
|
|
1332
|
+
t.Fatalf("subagent stop exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1333
|
+
}
|
|
1334
|
+
|
|
1335
|
+
stop := hookPayload(t, map[string]any{
|
|
1336
|
+
"hook_event_name": "Stop",
|
|
1337
|
+
"session_id": sessionID,
|
|
1338
|
+
"turn_id": turnID,
|
|
1339
|
+
"stop_hook_active": false,
|
|
1340
|
+
})
|
|
1341
|
+
if out, stderr, code := runDevritesIO(t, root, stop, nil,
|
|
1342
|
+
"hook", "agent-dispatch", "--harness=codex"); code != 0 || strings.TrimSpace(out) != "" {
|
|
1343
|
+
t.Fatalf("resolved stop exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1344
|
+
}
|
|
1345
|
+
}
|
|
1346
|
+
|
|
1347
|
+
func TestCodexAgentDispatchConfirmsDurableV2NamedRole(t *testing.T) {
|
|
1348
|
+
root := newWorkspace(t)
|
|
1349
|
+
codeHome := t.TempDir()
|
|
1350
|
+
sessionID, turnID := "session-v2-named", "turn-v2-named"
|
|
1351
|
+
role := "devrites-security-auditor"
|
|
1352
|
+
writeCodexAgentContract(t, root, role)
|
|
1353
|
+
writeCodexSkillContract(t, root, "devrites-audit", role)
|
|
1354
|
+
|
|
1355
|
+
runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1356
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1357
|
+
"session_id": sessionID,
|
|
1358
|
+
"turn_id": turnID,
|
|
1359
|
+
"prompt": "$devrites-audit Call spawn_agent and tell it to read .codex/agents/" + role + ".toml.",
|
|
1360
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1361
|
+
|
|
1362
|
+
writeCodexV2Rollouts(
|
|
1363
|
+
t, codeHome, filepath.Dir(root), sessionID, turnID, role,
|
|
1364
|
+
"devrites_security_auditor_named", true, true, true,
|
|
1365
|
+
)
|
|
1366
|
+
stop := hookPayload(t, map[string]any{
|
|
1367
|
+
"hook_event_name": "Stop",
|
|
1368
|
+
"session_id": sessionID,
|
|
1369
|
+
"turn_id": turnID,
|
|
1370
|
+
"stop_hook_active": false,
|
|
1371
|
+
})
|
|
1372
|
+
if out, stderr, code := runDevritesIO(t, root, stop,
|
|
1373
|
+
[]string{"CODEX_HOME=" + codeHome},
|
|
1374
|
+
"hook", "agent-dispatch", "--harness=codex"); code != 0 || strings.TrimSpace(out) != "" {
|
|
1375
|
+
t.Fatalf("durable v2 stop exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1376
|
+
}
|
|
1377
|
+
}
|
|
1378
|
+
|
|
1379
|
+
func TestCodexAgentDispatchRejectsIncompleteDurableV2(t *testing.T) {
|
|
1380
|
+
for _, tc := range []struct {
|
|
1381
|
+
name string
|
|
1382
|
+
role string
|
|
1383
|
+
includeWait bool
|
|
1384
|
+
includeRules bool
|
|
1385
|
+
includeResult bool
|
|
1386
|
+
}{
|
|
1387
|
+
{name: "generic-role", role: "explorer", includeWait: true, includeRules: true, includeResult: true},
|
|
1388
|
+
{name: "missing-wait", role: "devrites-security-auditor", includeRules: true, includeResult: true},
|
|
1389
|
+
{name: "missing-rules", role: "devrites-security-auditor", includeWait: true, includeResult: true},
|
|
1390
|
+
{name: "missing-result", role: "devrites-security-auditor", includeWait: true, includeRules: true},
|
|
1391
|
+
} {
|
|
1392
|
+
t.Run(tc.name, func(t *testing.T) {
|
|
1393
|
+
root := newWorkspace(t)
|
|
1394
|
+
codeHome := t.TempDir()
|
|
1395
|
+
sessionID, turnID := "session-"+tc.name, "turn-"+tc.name
|
|
1396
|
+
role := "devrites-security-auditor"
|
|
1397
|
+
writeCodexAgentContract(t, root, role)
|
|
1398
|
+
writeCodexSkillContract(t, root, "devrites-audit", role)
|
|
1399
|
+
runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1400
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1401
|
+
"session_id": sessionID,
|
|
1402
|
+
"turn_id": turnID,
|
|
1403
|
+
"prompt": "$devrites-audit Read .codex/agents/" + role + ".toml.",
|
|
1404
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1405
|
+
writeCodexV2Rollouts(
|
|
1406
|
+
t, codeHome, filepath.Dir(root), sessionID, turnID, tc.role,
|
|
1407
|
+
"devrites_security_auditor_named", tc.includeWait, tc.includeRules, tc.includeResult,
|
|
1408
|
+
)
|
|
1409
|
+
stop := hookPayload(t, map[string]any{
|
|
1410
|
+
"hook_event_name": "Stop",
|
|
1411
|
+
"session_id": sessionID,
|
|
1412
|
+
"turn_id": turnID,
|
|
1413
|
+
})
|
|
1414
|
+
out, stderr, code := runDevritesIO(t, root, stop,
|
|
1415
|
+
[]string{"CODEX_HOME=" + codeHome},
|
|
1416
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1417
|
+
if code != 0 {
|
|
1418
|
+
t.Fatalf("stop exit=%d stderr=%s", code, stderr)
|
|
1419
|
+
}
|
|
1420
|
+
var decision struct {
|
|
1421
|
+
Decision string `json:"decision"`
|
|
1422
|
+
}
|
|
1423
|
+
if err := json.Unmarshal([]byte(strings.TrimSpace(out)), &decision); err != nil {
|
|
1424
|
+
t.Fatalf("invalid Stop response: %v\n%s", err, out)
|
|
1425
|
+
}
|
|
1426
|
+
if decision.Decision != "block" {
|
|
1427
|
+
t.Fatalf("incomplete durable V2 dispatch passed: %s", out)
|
|
1428
|
+
}
|
|
1429
|
+
})
|
|
1430
|
+
}
|
|
1431
|
+
}
|
|
1432
|
+
|
|
1433
|
+
func TestCodexAgentDispatchGatesReconcileOnBoundWrightResult(t *testing.T) {
|
|
1434
|
+
root := newWorkspace(t)
|
|
1435
|
+
writeActive(t, root, "auth-tokens")
|
|
1436
|
+
workspace := filepath.Join(root, "features", "auth-tokens")
|
|
1437
|
+
for _, name := range []string{".reconcile-base", ".reconcile-allowlist", ".reconcile-devrites"} {
|
|
1438
|
+
if err := os.WriteFile(filepath.Join(workspace, name), []byte(name+"\n"), 0o600); err != nil {
|
|
1439
|
+
t.Fatal(err)
|
|
1440
|
+
}
|
|
1441
|
+
}
|
|
1442
|
+
sessionID, turnID := "session-wright", "turn-wright"
|
|
1443
|
+
writeCodexAgentContract(t, root, "devrites-slice-wright")
|
|
1444
|
+
writeCodexSkillContract(t, root, "rite-build", "devrites-slice-wright")
|
|
1445
|
+
runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1446
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1447
|
+
"session_id": sessionID,
|
|
1448
|
+
"turn_id": turnID,
|
|
1449
|
+
"prompt": "$rite-build",
|
|
1450
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1451
|
+
|
|
1452
|
+
reconcile := hookPayload(t, map[string]any{
|
|
1453
|
+
"hook_event_name": "PreToolUse",
|
|
1454
|
+
"session_id": sessionID,
|
|
1455
|
+
"turn_id": turnID,
|
|
1456
|
+
"tool_name": "Bash",
|
|
1457
|
+
"tool_input": map[string]any{"command": "rtk devrites-engine reconcile close"},
|
|
1458
|
+
})
|
|
1459
|
+
out, _, _ := runDevritesIO(t, root, reconcile, nil,
|
|
1460
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1461
|
+
if decision, reason := parsePermissionDecision(t, out); decision != "deny" ||
|
|
1462
|
+
!strings.Contains(reason, "confirmed") {
|
|
1463
|
+
t.Fatalf("premature reconcile not denied: decision=%q reason=%q", decision, reason)
|
|
1464
|
+
}
|
|
1465
|
+
|
|
1466
|
+
for _, payload := range []map[string]any{
|
|
1467
|
+
{
|
|
1468
|
+
"hook_event_name": "PreToolUse",
|
|
1469
|
+
"session_id": sessionID,
|
|
1470
|
+
"turn_id": turnID,
|
|
1471
|
+
"tool_name": "spawn_agent",
|
|
1472
|
+
"tool_use_id": "spawn-wright",
|
|
1473
|
+
"tool_input": map[string]any{
|
|
1474
|
+
"agent_type": "worker",
|
|
1475
|
+
"fork_turns": "none",
|
|
1476
|
+
"message": "Read .codex/agents/devrites-slice-wright.toml.",
|
|
1477
|
+
},
|
|
1478
|
+
},
|
|
1479
|
+
{
|
|
1480
|
+
"hook_event_name": "SubagentStart",
|
|
1481
|
+
"session_id": sessionID,
|
|
1482
|
+
"turn_id": "child-wright",
|
|
1483
|
+
"agent_id": "agent-wright",
|
|
1484
|
+
"agent_type": "worker",
|
|
1485
|
+
},
|
|
1486
|
+
{
|
|
1487
|
+
"hook_event_name": "SubagentStop",
|
|
1488
|
+
"session_id": sessionID,
|
|
1489
|
+
"turn_id": "child-wright",
|
|
1490
|
+
"agent_id": "agent-wright",
|
|
1491
|
+
"agent_type": "worker",
|
|
1492
|
+
"last_assistant_message": "agent-result/v1 complete",
|
|
1493
|
+
},
|
|
1494
|
+
{
|
|
1495
|
+
"hook_event_name": "PreToolUse",
|
|
1496
|
+
"session_id": sessionID,
|
|
1497
|
+
"turn_id": turnID,
|
|
1498
|
+
"tool_name": "wait",
|
|
1499
|
+
"tool_input": map[string]any{"ids": []string{"agent-wright"}},
|
|
1500
|
+
},
|
|
1501
|
+
} {
|
|
1502
|
+
hook := "agent-dispatch"
|
|
1503
|
+
if payload["hook_event_name"] == "SubagentStart" {
|
|
1504
|
+
hook = "subagent-orient"
|
|
1505
|
+
}
|
|
1506
|
+
if _, stderr, code := runDevritesIO(t, root, hookPayload(t, payload), nil,
|
|
1507
|
+
"hook", hook, "--harness=codex"); code != 0 {
|
|
1508
|
+
t.Fatalf("%s exit=%d stderr=%s", hook, code, stderr)
|
|
1509
|
+
}
|
|
1510
|
+
}
|
|
1511
|
+
|
|
1512
|
+
if out, stderr, code := runDevritesIO(t, root, reconcile, nil,
|
|
1513
|
+
"hook", "agent-dispatch", "--harness=codex"); code != 0 || strings.TrimSpace(out) != "" {
|
|
1514
|
+
t.Fatalf("completed wright reconcile exit=%d out=%s stderr=%s", code, out, stderr)
|
|
1515
|
+
}
|
|
1516
|
+
}
|
|
1517
|
+
|
|
1518
|
+
func TestCodexAgentDispatchBindsDurableV2WrightToReconcileWindow(t *testing.T) {
|
|
1519
|
+
for _, tc := range []struct {
|
|
1520
|
+
name string
|
|
1521
|
+
changeWindow bool
|
|
1522
|
+
wantPermitted bool
|
|
1523
|
+
}{
|
|
1524
|
+
{name: "same-window", wantPermitted: true},
|
|
1525
|
+
{name: "changed-window", changeWindow: true},
|
|
1526
|
+
} {
|
|
1527
|
+
t.Run(tc.name, func(t *testing.T) {
|
|
1528
|
+
root := newWorkspace(t)
|
|
1529
|
+
codeHome := t.TempDir()
|
|
1530
|
+
writeActive(t, root, "auth-tokens")
|
|
1531
|
+
workspace := filepath.Join(root, "features", "auth-tokens")
|
|
1532
|
+
for _, name := range []string{".reconcile-base", ".reconcile-allowlist", ".reconcile-devrites"} {
|
|
1533
|
+
if err := os.WriteFile(filepath.Join(workspace, name), []byte(name+"\n"), 0o600); err != nil {
|
|
1534
|
+
t.Fatal(err)
|
|
1535
|
+
}
|
|
1536
|
+
}
|
|
1537
|
+
sessionID, turnID := "session-v2-wright-"+tc.name, "turn-v2-wright-"+tc.name
|
|
1538
|
+
role := "devrites-slice-wright"
|
|
1539
|
+
writeCodexAgentContract(t, root, role)
|
|
1540
|
+
writeCodexSkillContract(t, root, "rite-build", role)
|
|
1541
|
+
runDevritesIO(t, root, hookPayload(t, map[string]any{
|
|
1542
|
+
"hook_event_name": "UserPromptSubmit",
|
|
1543
|
+
"session_id": sessionID,
|
|
1544
|
+
"turn_id": turnID,
|
|
1545
|
+
"prompt": "$rite-build",
|
|
1546
|
+
}), nil, "hook", "agent-dispatch", "--harness=codex")
|
|
1547
|
+
writeCodexV2Rollouts(
|
|
1548
|
+
t, codeHome, filepath.Dir(root), sessionID, turnID, role,
|
|
1549
|
+
"devrites_slice_wright_named", true, true, true,
|
|
1550
|
+
)
|
|
1551
|
+
if tc.changeWindow {
|
|
1552
|
+
future := time.Now().Add(time.Minute)
|
|
1553
|
+
if err := os.Chtimes(
|
|
1554
|
+
filepath.Join(workspace, ".reconcile-devrites"), future, future,
|
|
1555
|
+
); err != nil {
|
|
1556
|
+
t.Fatal(err)
|
|
1557
|
+
}
|
|
1558
|
+
}
|
|
1559
|
+
reconcile := hookPayload(t, map[string]any{
|
|
1560
|
+
"hook_event_name": "PreToolUse",
|
|
1561
|
+
"session_id": sessionID,
|
|
1562
|
+
"turn_id": turnID,
|
|
1563
|
+
"tool_name": "Bash",
|
|
1564
|
+
"tool_input": map[string]any{"command": "rtk devrites-engine reconcile close"},
|
|
1565
|
+
})
|
|
1566
|
+
out, stderr, code := runDevritesIO(t, root, reconcile,
|
|
1567
|
+
[]string{"CODEX_HOME=" + codeHome},
|
|
1568
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1569
|
+
if code != 0 {
|
|
1570
|
+
t.Fatalf("reconcile exit=%d stderr=%s", code, stderr)
|
|
1571
|
+
}
|
|
1572
|
+
if tc.wantPermitted && strings.TrimSpace(out) != "" {
|
|
1573
|
+
t.Fatalf("same-window V2 wright was rejected: %s", out)
|
|
1574
|
+
}
|
|
1575
|
+
if !tc.wantPermitted {
|
|
1576
|
+
if decision, _ := parsePermissionDecision(t, out); decision != "deny" {
|
|
1577
|
+
t.Fatalf("changed-window V2 wright was accepted: %s", out)
|
|
1578
|
+
}
|
|
1579
|
+
}
|
|
1580
|
+
})
|
|
1581
|
+
}
|
|
1582
|
+
}
|
|
1583
|
+
|
|
1584
|
+
func TestCodexAgentDispatchRejectsWrongGenericRole(t *testing.T) {
|
|
1585
|
+
root := newWorkspace(t)
|
|
1586
|
+
writeCodexAgentContract(t, root, "devrites-security-auditor")
|
|
1587
|
+
spawn := hookPayload(t, map[string]any{
|
|
1588
|
+
"hook_event_name": "PreToolUse",
|
|
1589
|
+
"session_id": "session-role",
|
|
1590
|
+
"turn_id": "turn-role",
|
|
1591
|
+
"tool_name": "spawn_agent",
|
|
1592
|
+
"tool_use_id": "spawn-role",
|
|
1593
|
+
"tool_input": map[string]any{
|
|
1594
|
+
"agent_type": "worker",
|
|
1595
|
+
"fork_turns": "none",
|
|
1596
|
+
"message": "Read .codex/agents/devrites-security-auditor.toml.",
|
|
1597
|
+
},
|
|
1598
|
+
})
|
|
1599
|
+
out, stderr, code := runDevritesIO(t, root, spawn, nil,
|
|
1600
|
+
"hook", "agent-dispatch", "--harness=codex")
|
|
1601
|
+
if code != 0 {
|
|
1602
|
+
t.Fatalf("exit=%d stderr=%s", code, stderr)
|
|
1603
|
+
}
|
|
1604
|
+
if decision, reason := parsePermissionDecision(t, out); decision != "deny" ||
|
|
1605
|
+
!strings.Contains(reason, "devrites-slice-wright") {
|
|
1606
|
+
t.Fatalf("worker role mismatch not denied: decision=%q reason=%q", decision, reason)
|
|
1607
|
+
}
|
|
1608
|
+
}
|