FA-11506 / Static analysis soundness / Open access

A context-specific sanitizer clears unrelated taint · case 01

A context-specific sanitizer clears unrelated taint.

Verified by executionVariant 1 · 6 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

A sanitizer clears all taint kinds rather than the kind it certifies.

THE FAILURE

A sanitizer clears all taint kinds rather than the kind it certifies.

Unsuccessful approach: Keeping only the certified kind inverts the certificate meaning and loses unrelated taint.

Case contract

Given hazard names and a sanitizer certificate name, remove exactly that name from the hazard set and return sorted remaining names. Unknown certificates remove nothing.

Why this case matters

A deterministic offline analysis model exposing a specific soundness or precision boundary; it does not implement a complete language analyzer.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(hazards, certificate):
    return [] if certificate else sorted(set(hazards))
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
h='sink'+str(N)
check('sql certificate preserves html',solve(['sql',h],'sql'),[h])
check('html certificate preserves sql',solve(['sql',h],h),['sql'])
check('unknown certificate',solve([h],'other'),[h])
check('matching certificate',solve([h],h),[])
check('empty hazards',solve([],h),[])
check('duplicate hazards',solve([h,h,'sql'],'sql'),[h])
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
sql certificate preserves html	`[]`	`['sink1']`	Failed
html certificate preserves sql	`[]`	`['sql']`	Failed
unknown certificate	`[]`	`['sink1']`	Failed
matching certificate	`[]`	`[]`	Passed
empty hazards	`[]`	`[]`	Passed
duplicate hazards	`[]`	`['sink1']`	Failed

SHA-256 / bb9f1c9525099ce7bceb723de8cb5d09eeb0e376b6109af3ad61737cde84cdcd

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(hazards, certificate):
    return sorted(set(hazards) & {certificate})
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
h='sink'+str(N)
check('sql certificate preserves html',solve(['sql',h],'sql'),[h])
check('html certificate preserves sql',solve(['sql',h],h),['sql'])
check('unknown certificate',solve([h],'other'),[h])
check('matching certificate',solve([h],h),[])
check('empty hazards',solve([],h),[])
check('duplicate hazards',solve([h,h,'sql'],'sql'),[h])
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
sql certificate preserves html	`['sql']`	`['sink1']`	Failed
html certificate preserves sql	`['sink1']`	`['sql']`	Failed
unknown certificate	`[]`	`['sink1']`	Failed
matching certificate	`['sink1']`	`[]`	Failed
empty hazards	`[]`	`[]`	Passed
duplicate hazards	`['sql']`	`['sink1']`	Failed

SHA-256 / 57cd59eb12beaffaebaee64db6c12c5843fe4c5b3afd34a834a4e5c3d2a9cff4

HELD IN THE MEMBER ARCHIVE

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

This mechanism has 6 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

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 2026-09-29T14:38:48.603094+00:00.

Case digest / cfbc4eabffd38afd78b1e1e9db9b227f2d579ad6f2199ee6590e811d92f83c3d