Files
benbrownandClaude Opus 4.8 899d55f4b6 gates: close the audit findings — enforce the gates the docs claimed
An adversarial audit found the enforcement story ahead of the code: gates
that were documented as hard gates were advisory, manual, or absent, and
several human-facing claims outran the machine record. This makes the
claims true.

High findings

- Differential fuzz was documentation, not enforcement: verdict hardcoded
  fuzz_iters=0 and diff-test hardcoded the gate to Skipped, while the case
  studies claimed 10k-iteration fuzz on the same wins. Fuzz is now a real
  pipeline gate (strict FUZZ-RESULT contract, executed count recorded), and
  it is inherently baseline-vs-candidate, so it runs on the accept path and
  reports Skipped in the standalone `just gates` flow. New hard rule: a
  machine `accepted` verdict is impossible unless the fuzz gate actually
  passed — otherwise it caps at needs-human-review.
- The agent's system prompt claimed FP/concurrency/unsafe patches were
  "routed to human review automatically". No such mechanism existed. Added
  a conservative lexical risk classifier (crates/verdict/src/risk.rs) that
  forces needs-human-review on any would-be accept carrying those signals;
  fp-tolerance equivalence always signals.
- The local verdict path ran agent-influenced code unsandboxed. It now runs
  under scripts/no-net.sh (network namespace) and fails closed. Upstream
  test suites are hash-verified against corpora/<t>/TESTSUITE.sha256 when
  present (scripts/pin-testsuite.sh generates them).

Also fixed, found while fixing the above

- The pipeline script's `set -e` was inert (brace group as the left operand
  of `&&`), so a failing candidate build did not abort the run and verdict
  could bench a stale binary. Now a scoped subshell whose status is branched
  on at top level.
- A wrapper that exits before exec'ing (no-net.sh refusing, bad wrapper path)
  stranded RUNNING.lock and left read_verdict polling forever. Failure
  accounting now lives outside the wrapper.
- harnessd's patch allowlist ignored git rename/copy headers and --no-prefix
  diffs; the rename-escape case is now a committed regression test.
- The workspace did not compile on Windows (ungated Unix APIs). cfg-gated;
  71 portable tests now run on both platforms.
- pin-check regenerated the corpus manifest instead of verifying it; it is
  now verify-only, with a deliberate `just pin-corpus` to re-pin.
- Max RSS is captured per run on Unix; PMU counters are opt-in diagnostics
  (compare --perf-stat), never part of the accept decision.

Docs now match the code, including what is NOT yet enforced: no target ships
a TESTSUITE.sha256 pin, no target declares [build].tsan, and matmul is not
wired to the fp-tolerance gate — each is stated as mechanized-but-inert
rather than done. Corrected the stale "33 ledger rows" (34), the playbook's
class-3 cJSON win (rejected), adversarial-review.md's sanitizer-separation
claim (five accepted rows carry sanitizers_clean=false), the unsupported
"40,000 requests" figure, and the blanket "historical rows record
fuzz_iters=0" (8 of 10; phase0-tokei-002/-003 record 4,332).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-14 13:10:14 -04:00

79 lines
2.9 KiB
Bash

#!/bin/sh
# Deterministically generate the cjson corpus — a large synthetic JSON
# document exercising every parse path: nested objects/arrays, strings
# with escapes and unicode, ints, floats (the strtod/sprintf hot path),
# booleans, nulls. No randomness — byte-identical on every machine, so
# the MANIFEST.sha256 pin is reproducible. Run from repo root:
# sh corpora/cjson/gen-corpus.sh [--check|--pin]
# Default/--check: regenerate and verify against MANIFEST.sha256 (never
# rewrites it). --pin: deliberately rewrite the manifest.
set -eu
mode="check"
case "${1:-}" in
"" | --check) ;;
--pin) mode="pin" ;;
*) echo "usage: gen-corpus.sh [--check|--pin]" >&2; exit 2 ;;
esac
DIR="$(cd "$(dirname "$0")" && pwd)"
OUT="$DIR/input/big.json"
mkdir -p "$DIR/input"
python3 - "$OUT" <<'PY'
import json, sys
# Deterministic: a fixed linear-congruential stream, no `random` module,
# so output is identical everywhere.
class Det:
def __init__(self, seed=2463534242):
self.s = seed
def next(self):
self.s ^= (self.s << 13) & 0xFFFFFFFF
self.s ^= (self.s >> 17)
self.s ^= (self.s << 5) & 0xFFFFFFFF
return self.s & 0xFFFFFFFF
def rng(self, n):
return self.next() % n
d = Det()
WORDS = ["alpha","bravo","charlie","delta","echo","foxtrot","golf","hotel",
"quote\"inside","tab\tchar","new\nline","back\\slash","unécode","☃snow"]
def record(i):
return {
"id": i,
"guid": f"{d.next():08x}-{d.next():04x}",
"active": bool(d.rng(2)),
"score": d.rng(1000000) / 1000.0, # float -> strtod/sprintf path
"ratio": d.rng(1 << 30) / (1 << 30), # dense fractional float
"count": d.rng(1 << 31) - (1 << 30), # signed int
"name": WORDS[d.rng(len(WORDS))] + str(d.rng(9999)),
"tags": [WORDS[d.rng(len(WORDS))] for _ in range(d.rng(6))],
"coords": [d.rng(1 << 20) / 997.0 for _ in range(3)],
"meta": None if d.rng(4) == 0 else {
"created": d.next(),
"nested": {"a": d.rng(100), "b": [d.rng(50) for _ in range(d.rng(4))]},
},
}
doc = {
"schema": "hotpath.cjson.corpus/v1",
"generated_by": "gen-corpus.sh (deterministic)",
"records": [record(i) for i in range(20000)],
}
with open(sys.argv[1], "w") as f:
json.dump(doc, f, ensure_ascii=False, separators=(",", ":"))
PY
echo "wrote $OUT ($(wc -c < "$OUT") bytes)"
if [ "$mode" = "pin" ]; then
( cd "$DIR/input" && find . -type f -print0 | sort -z | xargs -0 sha256sum ) > "$DIR/MANIFEST.sha256"
echo "re-pinned $DIR/MANIFEST.sha256 — a deliberate human action (corpora/README.md);"
echo "commit the new manifest with justification."
else
if [ ! -f "$DIR/MANIFEST.sha256" ]; then
echo "no MANIFEST.sha256 — pin the corpus with: sh corpora/cjson/gen-corpus.sh --pin" >&2
exit 1
fi
( cd "$DIR/input" && sha256sum -c ../MANIFEST.sha256 )
fi