FA-99891 / Agent workflow termination / Open access

Workflow: a changed state is mistaken for a repeated loop state · case 01

A healthy workflow is halted because one state field changed.

Verified by executionVariant 1 · 8 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

The detector treats any nonempty history as a repeated state, without comparing keys.

THE FAILURE

The detector treats any nonempty history as a repeated state, without comparing keys.

Unsuccessful approach: Treating a sufficiently long history as a cycle still halts on novel state keys.

Case contract

Return true iff current key occurs at least twice in the observed key sequence.

Why this case matters

This toy decision rule isolates one bounded agent-loop completion contract.

1 / The failure

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(seen, current):
    return bool(seen)
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('same state repeats', solve(['a','b'],'b'),True)
check('new state', solve(['a','b'],'c'),False)
check('empty history', solve([],'a'),False)
check('novel after one prior state', solve(['a'],'b'),False)
check('singleton match', solve(['x'],'x'),True)
check('different key', solve(['x'],'y'),False)
check('later repeated key', solve(['a','c','a'],'a'),True)
check('empty string key', solve([''],'x'),False)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
same state repeats	`True`	`True`	Passed
new state	`True`	`False`	Failed
empty history	`False`	`False`	Passed
novel after one prior state	`True`	`False`	Failed
singleton match	`True`	`True`	Passed
different key	`True`	`False`	Failed
later repeated key	`True`	`True`	Passed
empty string key	`True`	`False`	Failed

SHA-256 / 49446e8de1b172cb00371f7f12e21b599c967918cefdcde579d7970cfa37f177

2 / The unsuccessful fix

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(seen, current):
    return current in seen or len(seen) > 1
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('same state repeats', solve(['a','b'],'b'),True)
check('new state', solve(['a','b'],'c'),False)
check('empty history', solve([],'a'),False)
check('novel after one prior state', solve(['a'],'b'),False)
check('singleton match', solve(['x'],'x'),True)
check('different key', solve(['x'],'y'),False)
check('later repeated key', solve(['a','c','a'],'a'),True)
check('empty string key', solve([''],'x'),False)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
same state repeats	`True`	`True`	Passed
new state	`True`	`False`	Failed
empty history	`False`	`False`	Passed
novel after one prior state	`False`	`False`	Passed
singleton match	`True`	`True`	Passed
different key	`False`	`False`	Passed
later repeated key	`True`	`True`	Passed
empty string key	`False`	`False`	Passed

SHA-256 / 76ea6fa4be1ab147cb90d875c30518cf9f8398bae17c0fd3ba0918c05234fde4

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 8 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Verification & scope

Keys are canonical immutable strings; canonicalization itself is outside scope. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-30T16:48:29.419755+00:00.

Case digest / d774a86540698880001d0839cacbc8ed02cc70ac67c6b487f570aea07da18287