Skip to main content

AgentXchain v2.155.51

v2.155.51 fixes BUG-97: failed-acceptance reacceptance now normalizes stale retained-turn run_id values from the active state when the staged turn_id still matches the active turn. This closes the next DOGFOOD-100 recovery blocker discovered while resuming tusq.dev on the shipped package.

Bug Fixes​

  • BUG-97: normalize retained-turn run_id drift. If a staged result belongs to the active retained turn but carries a stale or missing top-level run_id, AgentXchain rewrites run_id from authoritative state.run_id before assignment validation.
  • Fail-closed boundary. The normalizer does not rewrite run_id when staged turn_id is missing or mismatched, or when active-turn run identity disagrees with state.run_id.
  • Prompt hardening. Dispatch prompts now tell agents to use only current assignment values for run_id, turn_id, role, and runtime_id, and never to copy run_id from old reports, history, previous dispatch bundles, or retained staging JSON.
  • Conformance alignment. TR-002 and TR-003 now encode safe active-turn run_id normalization as accepted behavior; negative regressions cover real ownership mismatches.

Evidence​

  • npm test -- --test-timeout=60000 -> 7288 tests / 1475 suites / 0 failures / 5 skipped

Tester Re-Run Contract​

Verify: npx --yes -p [email protected] -c "agentxchain --version"

For BUG-97: resume the retained tusq.dev failed-acceptance turn on the shipped package. The acceptance pipeline must auto-normalize a stale run_id only when the staged turn_id matches the active retained turn, then continue through assignment validation without operator-side accept-turn, unblock, or staging JSON edits.

For BUG-96 carry-forward: decision entries that include existing decision text but omit rationale must still auto-normalize before schema validation.

For BUG-95 carry-forward: staged results that use files_modified instead of files_changed and omit runtime_id, summary, artifact, and proposed_next_role must still auto-normalize before schema validation.

For review-only carry-forward: a review_only role with missing or empty objections must still fail protocol challenge enforcement.

Carry-forward tester checks for shipped-package closure proof:

  • BUG-54 startup watchdog evidence should quote startup_latency_ms, elapsed_since_spawn_ms, and runtimes.<id>.startup_watchdog_ms.
  • BUG-55/87 verification-output evidence should quote undeclared_verification_outputs, verification.produced_files, and git status --short after accept-turn followed by checkpoint-turn. Clean tree means BUG-55 is fixed for your reproduction.
  • BUG-52/53 continuous recovery evidence should quote phase_entered, reconciled_before_dispatch, and session_continuation.