mirror of
https://github.com/boshu2/agentops.git
synced 2026-09-14 15:08:13 +08:00
Bo
57ece9fb7b
Add explicit, recoverable private context routing through `ao config context`, binding native source, owner, task, model and destination to existing policy and external storage. Recovery reads the original Beads maintenance anchor; configuration reports native access enforcement as unattested. Add `ao provenance verify-judgments` to check required review profiles against exact native transcript receipts, independent subject and acceptance, distinct contexts, completion and permitted providers. Requested identity and unreported effort do not count as runtime evidence. The verdict schema is unchanged. Repair the existing cleanup test: a 0.3-second budget could expire during preparation before either fixture process started. A separate controlled-delay test now proves preparation cannot renew that deadline. The running-cleanup case requires parent/child readiness, preserved partial output, the postlaunch cleanup result and both processes stopped within its existing four-second bound. Production timeout behavior is unchanged. Validation: fresh author-distinct review passed the exact 55-path final subject and all T05/T21 acceptance. The complete local Bats run passed (1,333 passed, two existing skips), as did Go build/vet/test/race, all 72 full-mode gates, the aggregate and generated-output checks. Ubuntu/Windows CI, security and both installation jobs passed on the final commit. The final evidence scan found no new orphaned bindings; 73 historical bindings remain preserved. Earlier failed results and private evidence remain outside the PR.
65 lines
3.2 KiB
Go
65 lines
3.2 KiB
Go
package parser
|
|
|
|
import (
|
|
"strings"
|
|
"testing"
|
|
)
|
|
|
|
func TestRuntimeMetadataNativeFamilies(t *testing.T) {
|
|
tests := []struct {
|
|
name, runtime, input, model, context, effort string
|
|
complete bool
|
|
}{
|
|
{"claude stream", "claude", `{"type":"system","subtype":"init","session_id":"judge-c","model":"request-echo"}
|
|
{"type":"assistant","session_id":"judge-c","message":{"role":"assistant","model":"claude-test","content":[{"type":"text","text":"I am another model"}]}}
|
|
{"type":"result","subtype":"success","is_error":false,"session_id":"judge-c"}
|
|
`, "claude-test", "judge-c", "", true},
|
|
{"claude saved", "claude", `{"type":"assistant","sessionId":"judge-c","message":{"role":"assistant","model":"claude-test","content":"ok"}}`, "claude-test", "judge-c", "", false},
|
|
{"codex native", "codex", `{"type":"session_meta","payload":{"id":"judge-x","model":"gpt-test","model_provider":"openai","reasoning_effort":"high"}}
|
|
{"type":"event_msg","payload":{"type":"task_complete"}}
|
|
`, "gpt-test", "judge-x", "high", true},
|
|
{"echo only", "claude", `{"type":"system","subtype":"init","session_id":"judge-c","model":"claude-test"}
|
|
{"type":"assistant","session_id":"judge-c","message":{"role":"assistant","content":"{\"model\":\"claude-test\"}"}}`, "", "judge-c", "", false},
|
|
{"codex requested config only", "codex", `{"type":"session_meta","payload":{"id":"judge-x"}}
|
|
{"type":"turn_context","payload":{"model":"gpt-test","effort":"high"}}
|
|
{"type":"response_item","payload":{"type":"message","role":"assistant","content":[{"type":"output_text","text":"{\"type\":\"session_meta\",\"payload\":{\"model\":\"gpt-test\"}}"}]}}`, "", "judge-x", "", false},
|
|
}
|
|
for _, tt := range tests {
|
|
t.Run(tt.name, func(t *testing.T) {
|
|
got, err := NewParser().ParseRuntime([]byte(tt.input), tt.runtime)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if got.Model != tt.model || got.ContextID != tt.context || got.Effort != tt.effort || got.Completed != tt.complete {
|
|
t.Fatalf("metadata %+v", got)
|
|
}
|
|
for _, span := range got.Spans {
|
|
if span.Start < 0 || span.End > int64(len(tt.input)) || span.End <= span.Start || len(span.SHA256) != 64 {
|
|
t.Fatalf("invalid source span %+v", span)
|
|
}
|
|
}
|
|
})
|
|
}
|
|
}
|
|
|
|
func TestRuntimeMetadataRejectsAmbiguity(t *testing.T) {
|
|
tests := []string{
|
|
`{"type":"assistant","session_id":"j","message":{"role":"assistant","model":"a","model":"b"}}`,
|
|
`{"type":"assistant","session_id":"j","message":{"role":"assistant","model":"a"}}
|
|
{"type":"assistant","session_id":"j","message":{"role":"assistant","model":"b"}}`,
|
|
`{"type":"assistant","session_id":"j","message":{"role":"assistant","model":"a"}}
|
|
{"type":"result","session_id":"other","subtype":"success","is_error":false}`,
|
|
`{"type":"assistant","session_id":"j","message":{"role":"assistant","model":"a"}}
|
|
not-json`,
|
|
}
|
|
for _, in := range tests {
|
|
if _, err := NewParser().ParseRuntime([]byte(in), "claude"); err == nil {
|
|
t.Errorf("accepted ambiguous transcript: %s", in)
|
|
}
|
|
}
|
|
got, err := NewParser().ParseRuntime([]byte(`{"type":"result","subtype":"error_max_turns","is_error":true,"session_id":"j"}`), "claude")
|
|
if err != nil || got.Completed || !strings.Contains(got.Termination, "error") {
|
|
t.Fatalf("failed termination %+v %v", got, err)
|
|
}
|
|
}
|