Files
sanderling/internal/ltl/false_negative_test.go
T
pj 26b49b379a fix ltl semantics and unify action enumeration (#71)
* fix(ltl): give every thunk a construction identity

Two distinct unnamed predicates both described as "Thunk(...)", so obligation
collapse merged their residuals and could drop a live violation. Identity is
assigned at construction and the fields are unexported, so a thunk cannot be
built without one.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(ltl): reduce a thrown-predicate residual instead of panicking

The verifier substitutes an ErrorFormula for the residual of a property whose
predicate threw, and that residual is fed back in on the next step. reduce had
no case for it, so the run crashed. It re-reports the same failure now.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(ltl): make a bounded always the dual of a bounded eventually

G<=n(f) and not F<=n(not f) disagreed on traces where the inner was still
pending when the window closed, so nnf's negation normal form was not semantics
preserving. Both sides now range over the observations at which their inner can
definitely resolve: the eventually keeps a pending inner as a disjunct, and the
always discharges vacuously at window close.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(ltl): arm a one-shot root once per run

A root that carries its own horizon is one obligation for the whole run, not one
per observation. Re-instantiating a top-level eventually monitored G F<=n(p)
instead of F<=n(p) and left one live obligation per step behind; a bounded
always restarted its window every step and never closed.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(verifier): stop wrapping a top-level eventually in always

`eventually(p).within(300, "seconds")` as a property meant "within 300 seconds
of every step", which spawned an obligation per step with its own resolved
deadline. A 553-step run carried 553 of them and serialized a 75 KB residual.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(ltl): serialize the resolved deadline of a bounded window

Two obligations spawned at different steps from one duration-bounded formula
differ only in the deadline the evaluator resolved for them, so they serialized
identically and the trace erased a distinction the evaluator makes. The authored
window stays in amount/unit; the resolved deadline rides alongside.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(verifier): split a witness's origin step from its detection step

A deferred obligation spans two steps: the one that armed it and the one whose
reduction failed. They were conflated under one index, so the extractor snapshot
(which is the detecting step's state) was reported against the origin step.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(runner): record a witness's detection step in the trace

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* feat(replay-ui): show the step a violation was detected at

The witness evidence is the detecting step's state, so say which step that is
and let a reader jump to it.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(verifier): record the extractor state the predicates actually read

On the web path extractor bodies are evaluated in V8 and injected here, but only
the goja value was replaced. The trace diff and the violation witness therefore
described a state no property ever saw.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* refactor(spec): one candidate producer over one target-eligibility rule

Both hosts routed verbs themselves and both policies enumerated their own
actions, and all four drifted. Web sent `swipes` to scrollable containers only,
so swipe-to-dismiss on a list row was reachable on native and unreachable on
web; the model policy folded gestures its own way and could not reach what the
seeded picker drew.

A host now reports facts about every element and never decides which verb may
act on it: targets.ts acceptsTarget owns that for both. pick.ts builtinCandidates
is the single enumeration, and the model policy reads it through
__sanderlingEnumerateBuiltin__ instead of reimplementing it in Go.

Gesture verbs change with it: scrolls stay vertical over scrollable containers,
swipes go free-form in all four directions from any element with real bounds.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(runner): name a builtin scroll by its drag origin

A builtin gesture carries endpoints and no selector, so every scroll rendered as
"Scroll down " in the prompt's recent-action memory and two scrollable regions
were indistinguishable.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(chrome): clear storage over cdp instead of scripting an opaque origin

Launch runs while the tab is still on about:blank, whose opaque origin denies
storage access, so localStorage.clear() threw SecurityError and every web run
died at launch. Storage.clearDataForOrigin needs no navigation. The exception
helper lands here because "Uncaught" is what hid this for so long.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(chrome): enable the swiftshader webgl fallback

Headless Chrome runs with --disable-gpu, and without this flag it refuses the
software WebGL backend: getContext returns null, so a canvas-rendered app paints
nothing and every screenshot is identical black.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* fix(web): resolve testTag through data-testid or id

Compose Multiplatform emits its testTag into the element id, which the native
table already accepts via the resource-id alias. The two web selector tables
were the only place that rejected it.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* test(spec): type-check the spec api as part of make test

The fake runtime in api.test.ts did not return a chainable handle from extract,
so the file had not type-checked since named() was added. Wiring the check into
make test stops it drifting again.

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J

* docs(manual): one-shot eventually and the gesture verbs

Claude-Session: https://claude.ai/code/session_01Fj4wJUikdABuMQEETwW55J
2026-08-12 18:06:04 +05:30

111 lines
4.6 KiB
Go

package ltl
import (
"testing"
"time"
)
// thunkSeq returns a predicate that yields the given boolean per observation,
// repeating the last value once the sequence is exhausted.
func thunkSeq(values ...bool) func() (bool, error) {
step := 0
return func() (bool, error) {
value := values[len(values)-1]
if step < len(values) {
value = values[step]
}
step++
return value, nil
}
}
func runAndFinalize(formula Formula, steps int) (Verdict, *Evaluator) {
evaluator := NewEvaluator(formula)
var last Verdict
for index := range steps {
last = evaluator.ObserveAt(time.Unix(int64(index), 0))
if last == VerdictViolated {
return last, evaluator
}
}
return evaluator.Finalize(), evaluator
}
// Implies with a temporal antecedent must still evaluate the consequent at the
// current step. Always(p) holds over the observed run, so a false consequent
// Not(q) at the last step makes the implication violated. The pre-fix engine
// deferred the whole implication and reported Holds.
func TestImplies_TemporalAntecedent_ConsequentFalseViolates(t *testing.T) {
p := Thunk(thunkSeq(true, true, true))
q := Thunk(thunkSeq(false, false, true))
formula := Implies(Always(p), Not(q))
if verdict, _ := runAndFinalize(formula, 3); verdict != VerdictViolated {
t.Fatalf("Implies(Always(p),Not(q)) p=TTT q=FFT: got %v, want violated", verdict)
}
}
// Single-step witness of the same class: antecedent holds, consequent false.
func TestImplies_TemporalAntecedent_SingleStepViolates(t *testing.T) {
formula := Implies(Always(Thunk(thunkSeq(true))), Not(Thunk(thunkSeq(true))))
if verdict, _ := runAndFinalize(formula, 1); verdict != VerdictViolated {
t.Fatalf("Implies(Always(true),Not(true)): got %v, want violated", verdict)
}
}
// A consequent inside an Or must not mask the violation either.
func TestImplies_TemporalAntecedent_OrConsequentViolates(t *testing.T) {
formula := Implies(Always(Thunk(thunkSeq(true))), Or(Not(Thunk(thunkSeq(true))), Pure(false)))
if verdict, _ := runAndFinalize(formula, 1); verdict != VerdictViolated {
t.Fatalf("Implies(Always(true),Or(Not(true),false)): got %v, want violated", verdict)
}
}
// Control: a false antecedent makes the implication vacuously hold.
func TestImplies_TemporalAntecedent_FalseAntecedentHolds(t *testing.T) {
p := Thunk(thunkSeq(true, true, false))
q := Thunk(thunkSeq(false, false, true))
formula := Implies(Always(p), Not(q))
if verdict, _ := runAndFinalize(formula, 3); verdict != VerdictHolds {
t.Fatalf("Implies(Always(p),Not(q)) p=TTF q=FFT: got %v, want holds", verdict)
}
}
// A bounded Always whose inner is definitely false inside the window violates.
// The pre-fix window-close branch returned holds() unconditionally and lost
// this; the breach check runs before the window check and must stay there.
func TestBoundedAlways_ViolatedInnerInsideWindow(t *testing.T) {
inner := AlwaysFormula{Inner: Thunk(thunkSeq(false)), StepBound: 1, HasStepBound: true}
formula := Always(inner)
if verdict, _ := runAndFinalize(formula, 3); verdict != VerdictViolated {
t.Fatalf("Always(boundedAlways(false,1)): got %v, want violated", verdict)
}
}
// A bounded Always whose inner is still a deferred Next obligation when the
// window closes discharges vacuously: nothing was breached inside the window.
// This is the exact dual of the bounded Eventually violating when its inner has
// not held by the time the window closes
// (TestEventuallyWithinSteps_NextInnerHitsBoundFirstStep), and the pair is what
// makes nnf's G/F dualisation semantics preserving. It costs the deferred check
// the window closed on: G<=n and F<=n both range over the observations at which
// their inner can definitely resolve, never past them.
func TestBoundedAlways_PendingInnerDischargesAtWindowClose(t *testing.T) {
inner := AlwaysFormula{Inner: Next(Thunk(thunkSeq(false))), StepBound: 1, HasStepBound: true}
if verdict, _ := runAndFinalize(Always(inner), 3); verdict != VerdictHolds {
t.Fatalf("Always(boundedAlways(Next(false),1)): got %v, want holds", verdict)
}
if verdict, _ := runAndFinalize(Always(nnf(Not(inner))), 3); verdict != VerdictViolated {
t.Fatalf("its negation: got %v, want violated", verdict)
}
}
// A bounded Always whose inner genuinely holds each step stays satisfied: the
// fix must not turn a satisfied bounded window into a false positive.
func TestBoundedAlways_HoldingInnerStillHolds(t *testing.T) {
inner := AlwaysFormula{Inner: Not(Thunk(thunkSeq(false))), StepBound: 1, HasStepBound: true}
formula := Always(inner)
if verdict, _ := runAndFinalize(formula, 3); verdict != VerdictHolds {
t.Fatalf("Always(boundedAlways(Not(false),1)): got %v, want holds", verdict)
}
}