4.8 KiB
Session export handoff gate
Production model visits may publish their output to subsequent stages only after their exact recorded native session has settled and a complete, nonempty, nontruncated export file has passed normal-file and size verification. Ownership reads and final export attempts have bounded contexts independent of execution cancellation. Unconfirmed cleanup never exports or advances.
An export failure retains the computed step, outcome, target, output and file
evidence. The run parks nonterminally in waiting, with
export_pending_sequence identifying the computed visit and no finished_at.
This is not a human decision or a model failure. A separate
session_export_gates checkpoint is committed with the
step and its existing durable export job. Model retry is refused for that
checkpoint: recovery reads only the original session. Missing exact ownership
cannot be reconstructed or guessed and remains blocked.
The checkpoint reserves the workspace durably even though no agent is executing.
SessionExportReservations supplies those reservations to launch admission;
PendingSessionExportGate(runID, sequence) identifies the exact quiescent visit
eligible for export-only repair. Cancellation releases the reservation without
discarding or replaying the computation, and prevents automatic continuation.
The repair handler must permit waiting only when that predicate is true for
the requested sequence; unrelated human/resource waits remain ineligible.
Both CLI wait paths keep waiting for export recovery; run --wait still pauses
for genuine human decisions. Lifecycle events carry the same pending sequence,
and a lifecycle frame with zero/absent sequence clears it. Status-less repair
freshness frames leave it unchanged. Continuation is waiting -> running, never
terminal -> running. Startup normalizes historical failed/interrupted export
checkpoints back to waiting before clients attach.
After a successful automatic or manual repair, the worker re-verifies ownership
(including backend/reference) and the actual file, then rebuilds memory, turn
history and stage budgets from the recorded visits and follows the saved target.
Starting the successor consumes the gate in the same transaction as BeginStep;
terminal targets consume it with FinishRun. A restart before either transaction
can resume the saved target, but never replays the completed owner. Cancellation,
unconfirmed cleanup or an overlapping active workspace prevents continuation.
New repair files belong in the original private visit directory, not public handoff files. Historical public repair paths remain verifiable for compatibility. Only unowned injected test executors are exempt from native-session enforcement. An original computation failure routed directly to failure stays a failure even when its best-effort diagnostic export cannot be recovered.
Verification (2026-10-20)
Integrated changes use authenticated direct V2 API exports, tracked text-only model prompts, and native settled Pi JSONL export. V1 retains its native CLI export path. Native messages are retained as escaped JSON alongside the readable HTML projection so provider metadata and inline tool content are not dropped.
The original run #19 sessions were read without running any stage or changing
production evidence. Complete, nontruncated HTML exports in the approved
OpenCode temporary run19-export-proof/ directory contained:
| Stage | Messages | HTML bytes |
|---|---|---|
| Survey | 14 | 1,253,234 |
| Code | 11 | 325,930 |
| Review | 10 | 308,622 |
| UIReview | 16 | 1,358,507 |
| Land | 11 | 320,557 |
The old CLI silently cut exports at 262,144 bytes. These readbacks verify recovery
is possible for these exact original sessions; they are not production repair
records. Historical message-less /generate sessions cannot be reconstructed.
Checks passed after integration: full Go suite, build/vet, focused gate/recovery race tests, and browser audit fixtures. The browser fixture verifies export-only retry and refresh of an already-open Session tab. Large (>256 KiB) API fixtures, wrong identity, partial/empty exports, restart, cancellation, workspace reservation, and exactly-once successor admission have focused regressions. A full ordinary artifact inventory no longer blocks versioned audit repair. Repairs have a separate bounded allowance (twenty attempts, each at most 2 MiB), so neither the ordinary inventory's file count nor its byte budget consumes the mandatory audit-recovery capacity. The ordinary scan limits remain unchanged.
No shared service was restarted or redeployed. Production run #19's original evidence remains unchanged. A new real multi-stage run against the installed gate and native Linux execution remain unverified; source tests and read-only native readbacks do not substitute for those deployment checks.