Compare commits

..
Author SHA1 Message Date
Jason WoltjeandClaude Fable 5 1cc41f9813 fix(quality): enumeration guard sees naming, not comments — F1/F2/F3 from 20155
ci/woodpecker/pr/ci Pipeline was successful
F1 (demonstrated defect): surface 2 was a regex over raw YAML text, so a
commented-out ci.yml invocation still counted as enumerated — #1017's own
failure mode reproduced inside the fix, in the most common disable path
there is. Comment lines are now skipped before matching; needles n8/c4
pin both polarities (a comment is not enumeration; a comment naming an
absent path raises no false stale-enumeration). The residual trailing-
comment limit is documented where the grep lives.

F2 (structural): the guard was link [0] of the six-hop pnpm chain it
guards, so severing that chain silenced guard and guarded together.
ci.yml now invokes it directly from the sanitization step too — reachable
from both surfaces it audits, so either path alone keeps one instrument
running.

F3 (vocabulary): failure line and header said 'reachable'; F1/F2 prove
the guard measures NAMING. It now says 'enumerated' everywhere, matching
the success line.

Needles: 14 passed / 0 failed. Real tree: population 44, enumerated 25,
excluded (signed) 19, surfaces 38, unchanged — F2's line adds redundancy,
not membership. F1 demo replayed against the fix: commenting out
test-install-migration.sh now fails loud (UNENUMERATED, rc=1).

Written-by: pepper (sb-it-1-dt)
Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01NsKce8iZuSuRnu3gVMCBKB
2026-07-31 08:18:56 -05:00
Jason WoltjeandClaude Fable 5 f7dd57bc10 feat(quality): CI test-membership guard — the enumeration can no longer silently under-run the disk (#1017)
ci/woodpecker/pr/ci Pipeline was successful
CI reaches shell suites through two hand-enumerated surfaces (package.json
test:framework-shell; direct ci.yml commands), and a hand-enumerated allowlist
re-arms its own gap: a new suite never auto-joins, which is how 17 of 39
suite-shaped files under framework/tools were invisible to every CI run.

check-test-enumeration.sh, now the first link of test:framework-shell:
- parses BOTH surfaces (JSON + chain tokenization; ci.yml token extraction) —
  never hand-written line ranges, which three seats independently mis-scoped;
- checks BOTH directions: on-disk-but-unreachable fails, surface-named-but-
  absent fails (a rename manufactures a stale entry silently);
- population pattern is basename *test*.sh, deliberately broad and applied
  IDENTICALLY to both sides: the strict test-*.sh prefix cannot even name
  three real boundary files, one of which three independent censuses handled
  three different ways with no recorded judgement;
- the exclusions file gets the same rigor: no reason, stale path, contradictory
  entry, or out-of-population entry each FAIL. An exclusion is a recorded
  decision someone signed, not an omission nobody made.

The 17 previously-invisible suites enter as signed exclusions with honest
per-entry reasons (the #1007 five cite credential non-hermeticity; the push
guards cite one-run-green local measurements plus an unverified setsid
dependency; the unmeasured say unmeasured) — converting the silent gap into
attributable entries without trading it for a red seam. Burndown = make a
suite CI-reachable, delete its line; the guard enforces the bookkeeping.

Needle suite: 12 passed / 0 failed, both polarities per failure mode, asserted
on the guard's own words. Real-tree run at 826a8b3b: population 44 =
25 enumerated + 19 excluded, zero unsigned, all 38 surface-named paths present.

Written-by: pepper (sb-it-1-dt)
Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01NsKce8iZuSuRnu3gVMCBKB
2026-07-31 08:07:02 -05:00
4 changed files with 8 additions and 234 deletions
@@ -34,18 +34,12 @@ EOF
# get_remote_host and get_gitea_token are provided by detect-platform.sh
get_state_from_status_json() {
# Capture piped JSON BEFORE invoking `python3 - <<PY`. The heredoc binds
# stdin to the Python program text — so json.load(sys.stdin) inside would
# try to re-read stdin after `-` already consumed it for the program,
# yielding EOF and returning "unknown" every time. Pass payload via env.
local payload
payload=$(cat)
CI_QUEUE_STATUS_JSON="$payload" python3 - <<'PY'
python3 - <<'PY'
import json
import os
import sys
try:
payload = json.loads(os.environ.get("CI_QUEUE_STATUS_JSON", ""))
payload = json.load(sys.stdin)
except Exception:
print("unknown")
raise SystemExit(0)
@@ -89,15 +83,12 @@ PY
}
print_pending_contexts() {
# Same stdin hazard as get_state_from_status_json above — pass payload via env.
local payload
payload=$(cat)
CI_QUEUE_STATUS_JSON="$payload" python3 - <<'PY'
python3 - <<'PY'
import json
import os
import sys
try:
payload = json.loads(os.environ.get("CI_QUEUE_STATUS_JSON", ""))
payload = json.load(sys.stdin)
except Exception:
print("[ci-queue-wait] unable to decode status payload")
raise SystemExit(0)
@@ -1,205 +0,0 @@
#!/usr/bin/env bash
# Regression suite for #1019: ci-queue-wait.sh's status parser must return a
# state DERIVED FROM ITS INPUT.
#
# The defect this pins: both python sites piped a payload into `python3 - <<'PY'`.
# The heredoc binds stdin to the Python program text, so json.load(sys.stdin) saw
# EOF, the bare `except` fired, and the function returned "unknown" for every input
# — success, pending and failure alike. "unknown" then reaches an `exit 0` arm, so
# the gate-6 queue guard passed unconditionally on both the gitea and github paths.
#
# Every assertion below is on the RETURNED STATE STRING. A test that asserted only
# `rc=0` would have passed against the broken build, which is why this bug survived.
#
# The functions are extracted from the shipped wrapper rather than reimplemented, so
# this suite measures the code that actually runs.
set -uo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
# Overridable so a mutation test can point the suite at a deliberately-broken copy
# without touching the tracked file. The earlier version of this suite required
# `git stash` + `git checkout` to do that, which left a repo-global stash entry
# that any sibling worktree could have popped onto an unrelated branch.
WRAPPER="${CI_QUEUE_WAIT_WRAPPER:-$SCRIPT_DIR/ci-queue-wait.sh}"
PASS=0
FAIL=0
pass() { printf ' PASS %s\n' "$1"; PASS=$((PASS + 1)); }
fail() { printf ' FAIL %s\n' "$1"; FAIL=$((FAIL + 1)); }
if [[ ! -f "$WRAPPER" ]]; then
printf 'FATAL: wrapper not found at %s\n' "$WRAPPER" >&2
exit 1
fi
TMPDIR_T="$(mktemp -d)"
trap 'rm -rf "$TMPDIR_T"' EXIT
extract_fn() {
# $1 = function name, $2 = source file, $3 = destination
sed -n "/^$1()/,/^}/p" "$2" > "$3"
[[ -s "$3" ]] || { printf 'FATAL: could not extract %s from %s\n' "$1" "$2" >&2; exit 1; }
}
extract_fn get_state_from_status_json "$WRAPPER" "$TMPDIR_T/state.sh"
extract_fn print_pending_contexts "$WRAPPER" "$TMPDIR_T/contexts.sh"
state_of() {
# shellcheck disable=SC1091
( source "$TMPDIR_T/state.sh"; printf '%s' "$1" | get_state_from_status_json )
}
expect_state() {
local label="$1" payload="$2" want="$3" got
got="$(state_of "$payload")"
if [[ "$got" == "$want" ]]; then
pass "$label -> $want"
else
fail "$label -> got '$got', want '$want'"
fi
}
echo "=== parser returns a state derived from its input (#1019) ==="
expect_state "success payload" \
'{"state":"success","statuses":[{"status":"success","context":"ci/build"}]}' \
'terminal-success'
expect_state "pending payload" \
'{"state":"pending","statuses":[{"status":"pending","context":"ci/build"}]}' \
'pending'
expect_state "failure payload" \
'{"state":"failure","statuses":[{"status":"failure","context":"ci/build"}]}' \
'terminal-failure'
expect_state "mixed success+pending is pending" \
'{"state":"pending","statuses":[{"status":"success"},{"status":"pending"}]}' \
'pending'
expect_state "running counts as pending" \
'{"state":"pending","statuses":[{"status":"running"}]}' \
'pending'
expect_state "error counts as failure" \
'{"state":"failure","statuses":[{"status":"error"}]}' \
'terminal-failure'
expect_state "empty status set is no-status" \
'{"state":"","statuses":[]}' \
'no-status'
# Control. This is the ONE input for which "unknown" is correct. Without it, a
# regression that hardcoded "unknown" again would still fail the cases above but
# the suite would give no signal that "unknown" remains reachable when it should be.
expect_state "undecodable payload stays unknown" \
'not json at all' \
'unknown'
expect_state "unrecognised status vocabulary is unknown" \
'{"state":"weird","statuses":[{"status":"weird"}]}' \
'unknown'
echo "=== pending contexts are reported to the operator ==="
contexts_of() {
# shellcheck disable=SC1091
( source "$TMPDIR_T/contexts.sh"; printf '%s' "$1" | print_pending_contexts )
}
out="$(contexts_of '{"statuses":[{"status":"pending","context":"ci/alpha"},{"status":"pending","context":"ci/beta"}]}')"
if grep -q 'ci/alpha' <<<"$out" && grep -q 'ci/beta' <<<"$out"; then
pass "both pending contexts emitted"
else
fail "pending contexts not emitted; got: $out"
fi
out="$(contexts_of '{"statuses":[{"status":"success","context":"ci/alpha"}]}')"
# Assert the POSITIVE message, not merely the absence of the context name. Absence
# alone is satisfied by total silence — and the pre-fix build was silent, so an
# absence-only assertion passed against the very defect this suite exists to catch.
if grep -q 'ci/alpha' <<<"$out"; then
fail "a non-pending context was emitted; got: $out"
elif grep -q 'no pending contexts' <<<"$out"; then
pass "non-pending context suppressed, and reported as 'no pending contexts'"
else
fail "expected an explicit 'no pending contexts' report; got: $out"
fi
echo "=== needle: the broken construct is caught, not merely absent today ==="
# Rebuild the pre-fix form and assert this suite would have failed against it.
# Without this, the suite proves the current file is correct but not that it can
# detect the defect returning.
BROKEN="$TMPDIR_T/broken.sh"
cat > "$BROKEN" <<'BROKEN_EOF'
get_state_from_status_json() {
python3 - <<'PY'
import json
import sys
try:
payload = json.load(sys.stdin)
except Exception:
print("unknown")
raise SystemExit(0)
print("terminal-success" if (payload.get("state") or "") == "success" else "pending")
PY
}
BROKEN_EOF
broken_got="$( ( source "$BROKEN"; printf '%s' '{"state":"success","statuses":[]}' | get_state_from_status_json ) )"
if [[ "$broken_got" == "unknown" ]]; then
pass "[NEEDLE ] pre-fix construct reproduces the defect (returns 'unknown' for a success payload)"
else
fail "[NEEDLE ] pre-fix construct did NOT reproduce the defect; got '$broken_got' — the needle no longer pins anything"
fi
# And assert the shipped wrapper does not contain that construct.
#
# The check must have HEREDOC SEMANTICS, not merely match the text. `json.load(sys.stdin)`
# is perfectly correct under `python3 -c '...'` — there the program comes from argv, so
# stdin really is the payload, and ci-queue-wait.sh uses that form legitimately in
# gitea_get_branch_head_sha. Only `python3 - <<DELIM` is defective, because `-` has
# already bound stdin to the program text.
#
# A flat `grep json\.load\(sys\.stdin\)` therefore fails on correct code. It did: the
# first version of this needle flagged the innocent `python3 -c` site. Same shape as
# #1018 F1 (a regex over raw text has no syntax semantics) reproduced inside the needle
# written to pin a different instance of it.
heredoc_stdin_sites() {
awk '
/python3[[:space:]]+-[[:space:]]*<</ {
d = $0
sub(/.*<<[[:space:]]*/, "", d)
gsub(/['"'"'"]/, "", d)
delim = d; inhd = 1; next
}
inhd && $0 == delim { inhd = 0; next }
inhd {
line = $0
sub(/[[:space:]]*#.*$/, "", line)
if (line ~ /sys\.stdin/) printf "%d: %s\n", FNR, $0
}
' "$1"
}
# Positive control FIRST. An empty result from a broken scanner looks exactly like a
# clean file, so the scanner is proven to fire before its silence is trusted.
if [[ -n "$(heredoc_stdin_sites "$BROKEN")" ]]; then
pass "[NEEDLE ] scanner detects a stdin read inside a python3-heredoc"
else
fail "[NEEDLE ] scanner did NOT fire on the known-broken form — its silence proves nothing"
fi
sites="$(heredoc_stdin_sites "$WRAPPER")"
if [[ -n "$sites" ]]; then
fail "[NEEDLE ] wrapper reads stdin inside a python3-heredoc at: $sites"
else
pass "[NEEDLE ] wrapper has no stdin read inside any python3-heredoc"
fi
printf '\nci-queue-wait parse: %d passed, %d failed\n' "$PASS" "$FAIL"
[[ "$FAIL" -eq 0 ]] || exit 1
@@ -529,19 +529,7 @@ cmd_run() {
echo "detector.sh: WARN — off-host liveness beacon emit failed (see beacon.sh); the off-host absence check remains the authoritative dead-man." >&2
fi
[ "$once" -eq 1 ] && break
# Close the detector lock fd in the sleep child; otherwise an orphaned sleep
# keeps the single-instance flock (fd 9, taken at exec 9> above) alive after
# the detector parent dies. The lock is non-blocking (`flock -n`, above), so
# for as long as that sleep survives a replacement instance is REFUSED and
# exits rather than queueing. This particular hold is BOUNDED by one poll
# interval (WAKE_DETECTOR_INTERVAL, default 30s): when the orphaned sleep
# exits its copy of fd 9 closes, ending this bounded sleep-child hold. It
# does NOT follow that the next start succeeds — other inheritors of fd 9
# (the M1 adapter, M2 sink grandchildren) are outside this patch's scope and
# can keep holding the flock. The cost this removes is a restart window in
# which every supervisor retry fails on the sleep child's account.
# `9>&-` closes ONLY the child's copy — the parent's lock is unaffected.
sleep "$interval" 9>&-
sleep "$interval"
done
}
+1 -1
View File
@@ -25,7 +25,7 @@
"lint": "eslint src",
"typecheck": "tsc --noEmit",
"test": "vitest run --passWithNoTests && pnpm run test:framework-shell",
"test:framework-shell": "bash framework/tools/quality/scripts/check-test-enumeration.sh && bash framework/tools/quality/scripts/test-check-test-enumeration.sh && python3 src/lease-broker/daemon_deadline_unittest.py && python3 src/lease-broker/normative_fragments_unittest.py && python3 src/lease-broker/receipt_challenge_unittest.py && python3 src/lease-broker/context_recovery_unittest.py && python3 src/lease-broker/recovery_runtime_unittest.py && python3 src/lease-broker/recovery_b1_adversarial_unittest.py && python3 src/lease-broker/framework_skill_portability_unittest.py && python3 src/mutator-gate/runtime_tools_unittest.py && python3 src/mutator-gate/runtime_launch_guard_unittest.py && python3 src/mutator-gate/version_coupling_unittest.py && python3 framework/tools/lease-broker/check-runtime-launches.py --root ../.. && bash framework/tools/codex/test-pr-diff-context.sh && bash framework/tools/qa/test-deps-preflight.sh && bash framework/tools/git/test-pr-review-gitea-comment.sh && bash framework/tools/git/test-pr-review-repo-host-override.sh && bash framework/tools/git/test-ci-queue-wait-branch-absent.sh && bash framework/tools/git/test-ci-queue-wait-parse.sh && bash framework/tools/git/test-git-credential-mosaic.sh && bash framework/tools/git/test-gitea-token-identity.sh && bash framework/tools/_scripts/test-install-ordering-guard.sh && bash framework/tools/tmux/agent-send.test.sh && bash framework/tools/wake/test-wake-store-ack.sh && bash framework/tools/wake/test-wake-store-enqueue-race.sh && bash framework/tools/wake/test-wake-digest-hmac.sh && bash framework/tools/wake/test-wake-digest-quarantine.sh && bash framework/tools/wake/test-wake-detector.sh && bash framework/tools/wake/test-wake-fn-oracle.sh && bash framework/tools/wake/test-wake-reconcile.sh && bash framework/tools/wake/test-wake-beacon.sh && bash framework/tools/wake/test-wake-preimage.sh && bash framework/tools/wake/test-wake-install.sh"
"test:framework-shell": "bash framework/tools/quality/scripts/check-test-enumeration.sh && bash framework/tools/quality/scripts/test-check-test-enumeration.sh && python3 src/lease-broker/daemon_deadline_unittest.py && python3 src/lease-broker/normative_fragments_unittest.py && python3 src/lease-broker/receipt_challenge_unittest.py && python3 src/lease-broker/context_recovery_unittest.py && python3 src/lease-broker/recovery_runtime_unittest.py && python3 src/lease-broker/recovery_b1_adversarial_unittest.py && python3 src/lease-broker/framework_skill_portability_unittest.py && python3 src/mutator-gate/runtime_tools_unittest.py && python3 src/mutator-gate/runtime_launch_guard_unittest.py && python3 src/mutator-gate/version_coupling_unittest.py && python3 framework/tools/lease-broker/check-runtime-launches.py --root ../.. && bash framework/tools/codex/test-pr-diff-context.sh && bash framework/tools/qa/test-deps-preflight.sh && bash framework/tools/git/test-pr-review-gitea-comment.sh && bash framework/tools/git/test-pr-review-repo-host-override.sh && bash framework/tools/git/test-ci-queue-wait-branch-absent.sh && bash framework/tools/git/test-git-credential-mosaic.sh && bash framework/tools/git/test-gitea-token-identity.sh && bash framework/tools/_scripts/test-install-ordering-guard.sh && bash framework/tools/tmux/agent-send.test.sh && bash framework/tools/wake/test-wake-store-ack.sh && bash framework/tools/wake/test-wake-store-enqueue-race.sh && bash framework/tools/wake/test-wake-digest-hmac.sh && bash framework/tools/wake/test-wake-digest-quarantine.sh && bash framework/tools/wake/test-wake-detector.sh && bash framework/tools/wake/test-wake-fn-oracle.sh && bash framework/tools/wake/test-wake-reconcile.sh && bash framework/tools/wake/test-wake-beacon.sh && bash framework/tools/wake/test-wake-preimage.sh && bash framework/tools/wake/test-wake-install.sh"
},
"dependencies": {
"@mosaicstack/brain": "workspace:*",