fix(orchestrator): task-scoped oscillation breaker for escalate/unblock ping-pong (#685)

* fix(orchestrator): task-scoped oscillation breaker for escalate/unblock ping-pong

An escalation ping-pong oscillates a task between two agents (cell PM
escalate_up -> BLOCKED -> main PM unblock -> restored -> respawn ->
escalate again). The per-(agent, task) respawn gate never trips on it:
the restored side is dispatched by _dispatch_claimed_without_agent,
which consults no respawn counter at all, so one side of the round trip
always has fuel regardless of the other's strikes — and even a tripped
main-PM counter only stalls the task silently at blocked instead of
surfacing the oscillation.

- Strikes are counted task-scoped at the unblock() chokepoint
  (agent-agnostic; legitimate needs_revision rework never calls
  unblock, so it structurally cannot trip this), durable in the
  existing orchestration_markers column — no migration.
- Progress between round-trips (commits / revision_count advancing)
  resets the count: real forward motion is not an oscillation.
- On trip: the task is blocked with a HUMAN resolver (the budget-breach
  posture), both dispatchers stop respawning onto it, further unblock()
  refuses until an admin override clears the marker, and the CEO
  notification names both agents and the cycle count.
- _notification_has_live_work now treats a HITL-blocked related task as
  no live work, closing the same loop for the admin-route escalation
  path.

* fix(orchestrator): wire the oscillation trip to the dispatchers and make recovery reachable

- TaskResponse serializes blocker_resolver_type: the dispatchers' HITL-blocked
  skip and the notification-path live-work check now actually fire over the
  wire instead of only against in-process rows.
- The oscillation marker clears on every human transition out of BLOCKED
  (snapshot or not), and the human unblock route treats a tripped task as
  the requested intervention: clears the marker and proceeds, while the
  agent gateway verb keeps refusing.
- The progress fingerprint includes the terminal-children count, so a
  coordination root whose children advanced between escalations resets
  instead of accruing toward a false trip.

---------

Co-authored-by: Renn F <rennf93@users.noreply.github.com>
This commit is contained in:
Renzo F
2026-07-24 17:20:20 +02:00
committed by GitHub
co-authored by Renn F
parent 23ae0ca217
commit eb0dcb6ecb
11 changed files with 990 additions and 66 deletions
@@ -8,6 +8,7 @@ from roboco.foundation.policy.content import markers as m
# Named constant — ruff PLR2004 forbids magic-value comparisons.
_TWO = 2
_THREE = 3
def _task(om: dict | None = None) -> SimpleNamespace:
@@ -155,3 +156,47 @@ def test_block_flip_count_bump_and_notify() -> None:
assert m.is_block_flip_notified(t) is True
# Marking notified must not reset the counter.
assert m.get_block_flip_count(t) == _TWO
def test_oscillation_strikes_accrue_on_unchanged_fingerprint() -> None:
t = _task()
assert m.get_oscillation_strikes(t) == 0
assert m.is_oscillation_tripped(t) is False
assert m.bump_oscillation_strikes(t, [0, 0]) == 1
assert m.bump_oscillation_strikes(t, [0, 0]) == _TWO
assert m.bump_oscillation_strikes(t, [0, 0]) == _THREE
assert m.get_oscillation_strikes(t) == _THREE
assert m.is_oscillation_tripped(t) is False
def test_oscillation_strikes_reset_on_progress() -> None:
t = _task()
m.bump_oscillation_strikes(t, [0, 0])
assert m.bump_oscillation_strikes(t, [0, 0]) == _TWO
# A new commit landed between rounds — real progress resets to 1.
assert m.bump_oscillation_strikes(t, [1, 0]) == 1
# A revision round completing is progress too.
assert m.bump_oscillation_strikes(t, [1, 0]) == _TWO
assert m.bump_oscillation_strikes(t, [1, 1]) == 1
def test_mark_oscillation_tripped_preserves_strikes_and_fingerprint() -> None:
t = _task()
m.bump_oscillation_strikes(t, [2, 1])
m.bump_oscillation_strikes(t, [2, 1])
m.mark_oscillation_tripped(t)
assert m.is_oscillation_tripped(t) is True
assert m.get_oscillation_strikes(t) == _TWO
# tripped survives a subsequent bump (belt-and-suspenders — the guard is
# meant to refuse before another bump ever happens).
m.bump_oscillation_strikes(t, [2, 1])
assert m.is_oscillation_tripped(t) is True
def test_clear_marker_removes_oscillation_state() -> None:
t = _task()
m.bump_oscillation_strikes(t, [0, 0])
m.mark_oscillation_tripped(t)
m.clear_marker(t, m.OSCILLATION_STRIKES)
assert m.get_oscillation_strikes(t) == 0
assert m.is_oscillation_tripped(t) is False