Skip to content

fix(limiter): complete pre-auth family coverage + advisory remediation (next 16.4.6, body-parser, zod) - #462

Closed
Alajmah wants to merge 1 commit into
masterfrom
security/limiter-identity-v7
Closed

Alajmah wants to merge 1 commit into
masterfrom
security/limiter-identity-v7

Conversation

@Alajmah

@Alajmah Alajmah commented Nov 20, 2026

Copy link
Copy Markdown
Contributor

Final branch of the limiter identity remediation (supersedes #454 → #456 → #458 → #459 → #460 → #461; branch updates ruleset-refused). Carries the pre-auth family extension AND the advisory remediation the #457 master-push CI failed on.

1. Pre-auth family extension (the required correction)

/api/auth/check (session-cookie Redis lookup) and /api/bootstrap/state + /first (anonymous DB work) both mount before apiKeyAuth and carried the same bucket-rotation mechanism. The pre-auth match now covers the complete /api/auth/* and /api/bootstrap/* families — prefix regex, case + trailing-slash tolerant. Bearer identity unchanged for ordinary routes.

New regressions (3): AUTH-CHECK end-to-end (130 rotating-Bearer session requests → 429 at 121 on ONE bucket); BOOTSTRAP-STATE end-to-end (130 anonymous → 429 at 121 on ONE bucket); FAMILY VARIANTS (8 path forms all key on the IP). All prior regressions carry.

2. Advisory remediation (the #457 master-push CI failure)

Seven newly published advisories (three critical) on next ≤16.4.0, body-parser <1.20.9, and zod <4.6.6. This branch carries the bumps AND the build fixes the version bump required:

  • next 16.3.8 → 16.4.6 (dashboard) — needs middleware.ts → proxy.ts (file AND function export renamed) + @next/swc-* platform binaries hand-carried to 16.4.6 (the version mismatch crashed builds with Cannot read properties of undefined (reading 'project')).
  • next 15.5.24 → 15.5.28 (landing; the same advisories' 13.x–15.x range).
  • body-parser 1.20.6 → 1.20.9 (override); zod 3.25.76 → 4.6.7 (web).
  • Lockfile entries hand-carried with registry integrities (the known Windows npm quirk with overrides + platform binaries).

Production is still at 8c9b93b (the #457 deploy correctly skipped on the failed audit). Merging this branch restores the audit gate and deploys the limiter fix, the #457 body cap, and the advisory remediation together.

Verification

  • Audit gate: 0 high/critical (local).
  • Limiter identity (17) + behavior (9) + webhook body-limit (14) = 40 tests green.
  • Dashboard: 5 suites / 66 tests green; build verified locally (compiled successfully, 29/29 static pages generated).

Holding at the merge gate — exact head b1aca04ebb071ba36ded730b754b880e5fccaa07.

…n (next 16.4.6, body-parser, zod)

PRE-AUTH EXTENSION (maintainer finding on #460): /api/auth/check
(session-cookie Redis lookup) and /api/bootstrap/state + /first
(anonymous DB work) mount before apiKeyAuth and carried the same
bucket-rotation mechanism. The pre-auth match now covers the COMPLETE
/api/auth/* and /api/bootstrap/* families (prefix regex, case +
trailing-slash tolerant). Bearer identity unchanged for ordinary
authenticated routes. Three new regressions: AUTH-CHECK end-to-end
(130 rotating-Bearer session requests → 429@121 ONE bucket);
BOOTSTRAP-STATE end-to-end (130 anonymous → 429@121 ONE bucket);
FAMILY VARIANTS (8 path forms all key on the IP).

THREE-LAYER REMEDIATION (carried): session-cookie rotation closed
(apiKeyAuth-exact recognition); pre-auth login closed (IP-keyed);
route-variant closed (trailing slash + case); derivation isolated
(derived as an exported function — the CodeQL missing-rate-limiting
model pairing resolved without behavior change).

ADVISORY REMEDIATION (the #457 master-push CI failed on seven new
advisories, three critical, on next ≤16.4.0):
- next 16.3.8→16.4.6 (dashboard; GHSA-2464 critical, GHSA-673q
  critical, GHSA-p3m5 high): requires middleware.ts→proxy.ts rename
  (function export renamed too) + @next/swc platform binaries
  hand-carried to 16.4.6 (the version mismatch crashed builds with
  'Cannot read properties of undefined (reading project)').
- next 15.5.24→15.5.28 (landing; same advisories' 13.x-15.x range).
- body-parser 1.20.6→1.20.9 (override; GHSA-27r3 high).
- zod 3.25.76→4.6.7 (web; GHSA-mrhx high).
Lockfile entries hand-carried with registry integrities (npm on
Windows refused to re-resolve overrides + platform binaries — the
known quirk).

VERIFICATION: audit gate 0 high/critical; limiter identity (17) +
behavior (9) + webhook body-limit (14) = 40 tests green; dashboard
5 suites / 66 tests green; dashboard build verified locally
(compiled successfully, 29/29 static pages).

Signed-off-by: Alajmah <55954113+Alajmah@users.noreply.github.com>
@Alajmah

Alajmah commented Nov 20, 2026

Copy link
Copy Markdown
Contributor Author

/gitwire run review

@chatgpt-codex-connector

Copy link
Copy Markdown
Codex Review Summary: ❌ Failed

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ❌ Failed 2026-11-20T19:41:35.056498Z
We couldn’t complete this review. Comment @codex review to try again.
b1aca04 PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector

Copy link
Copy Markdown

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

@gitwire-hq

gitwire-hq Bot commented Nov 20, 2026

Copy link
Copy Markdown
Contributor

▶️ GitWire: Re-evaluation triggered for review. Results will appear shortly.

@gitwire-hq gitwire-hq Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🤖 AI Code Review — 💬 Needs discussion

Confidence: low · Findings: 6
AI judgment: NEEDS DISCUSSION · Evidence: COMPLETE · Authority: advisory
Evidence complete · 8/8 changed files accounted for

Other findings (6)

  • [medium] zod major bump (3 → 4) with no accompanying source changes — verify v4 compatibility
  • [low] Misleading test comment cites a non-existent regex for the tab-scheme case
  • [medium] Auth/limiter prefix divergence: 'Bearertoken' may authenticate via apiKeyAuth's regex yet be bucketed as anonymous IP ⚠️ unverified
  • [low] apiKeyAuth ↔ deriveLimiterIdentity lockstep is enforced by a comment, not by code ⚠️ unverified
  • [medium] Next.js and body-parser bumps received none of the compatibility scrutiny applied to zod ⚠️ unverified

🔮 Devil's Advocate (2 turns): 4 missed risks found · 2 upheld


GitWire AI Review Gate (bundle-driven v2) · Structured schema · Scope-validated · Devil's Advocate · AI recommendation — maintainers and repository policy retain authority

Finding details (not annotatable inline)

  • [MEDIUM] Auth/limiter prefix divergence: 'Bearertoken' may authenticate via apiKeyAuth's regex yet be bucketed as anonymous IP

    The test comment implies apiKeyAuth matches ^Bearer\s+ (whitespace-tolerant) while deriveLimiterIdentity uses a strict slice(0,7) === 'Bearer ' comparison. If both hold, a header like 'Bearer\tvalid-token' passes authentication but lands in the pre-auth/IP bucket: authenticated users get throttled as anonymous, share an IP bucket with other NAT users, or an attacker can deliberately route authenticated traffic through the looser path. For a PR whose stated purpose is 'complete pre-auth family coverage', this contract seam between the auth parser and the identity function is squarely in scope and was never examined.

  • [LOW] apiKeyAuth ↔ deriveLimiterIdentity lockstep is enforced by a comment, not by code

    The new test documents that the slice(0,7) prefix 'must stay in lockstep with apiKeyAuth', but nothing enforces it: no shared exported constant (e.g., BEARER_PREFIX) and no contract test asserting both mechanisms agree on the same header corpus. The pre-auth coverage bug this PR fixes was plausibly caused by exactly this kind of drift, and without structural or test-level enforcement it can be silently reintroduced by an edit to either side.

  • [MEDIUM] Next.js and body-parser bumps received none of the compatibility scrutiny applied to zod

    The review audited the zod major bump in detail but said nothing about the next→16.4.6 and body-parser bumps in the same remediation commit. If the next bump crossed a major (15→16), it carries its own breakage surface — Node/React version floors, changed default caching/middleware behavior, typically requiring codemods; even a patch-level remediation bump warrants confirming the React peer range and build output. A body-parser bump can also drift from the pinned express version's expectations. Uneven scrutiny: one dependency got a compatibility question, the others got zero.

  • [LOW] Extending limiter coverage to previously-unlimited pre-auth routes can throttle legitimate unauthenticated traffic

    'Complete pre-auth family coverage' implies routes that previously had no limit now get one. Health checks, CI pollers, webhooks, and shared-NAT/mobile clients can newly trip the anonymous/IP bucket and start receiving 429s. The review verified the identity derivation logic but not the bucket sizes/windows applied to the newly covered routes — an operational regression a green test suite will not catch.

Comment thread packages/web/package.json
"pino-pretty": "^10.0.0",
"yaml": "^2.9.0",
"zod": "^3.22.0"
"zod": "^4.6.6"

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[MEDIUM] zod major bump (3 → 4) with no accompanying source changes — verify v4 compatibility

packages/web/package.json jumps zod from ^3.22.0 to ^4.6.6. That is a major-version upgrade and Zod 4 ships real breaking changes: z.record() now requires both key and value schemas (single-arg form throws at schema-definition time), the message: customization parameter was renamed to error on many APIs, error formatting changed, and some string-format methods are deprecated/removed. No source changes in packages/web accompany this bump, so either the existing schemas are already v4-safe, or validation code breaks/alters behavior at runtime even though a build/audit gate passes. Suggested fix: grep packages/web/src for single-argument z.record( calls and message: options in schemas, and run the schema-dependent integration tests against v4 before merging. If usage is confirmed compatible, consider pinning exactly 4.6.6+ (matching the advisory floor) rather than a caret range that floats across future majors.

// A LOWERCASE scheme is not Bearer for apiKeyAuth (starts-with "Bearer "),
// so it must not be Bearer for the limiter either — it falls to the IP.
redisState.counts.clear();
await runMiddleware({ headers: { authorization: "bearer key-lower" }, ip: "10.0.0.2" });

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[LOW] Misleading test comment cites a non-existent regex for the tab-scheme case

In the it.each block of rate-limiter-identity.test.js, the comment for ["Bearer\tgap-token"] says 'tab not matched by \s+? (regex ^Bearer\s+ — tab IS \s; assert below)', but deriveLimiterIdentity contains no regex. The actual mechanism is the exact 7-character prefix comparison authHeader.slice(0, 7) === "Bearer " (a tab at position 7 fails that strict equality). The assertion itself passes for the correct reason — the header falls back to the IP bucket — but the comment describes the wrong mechanism and could mislead a future maintainer who edits the identity contract or 'fixes' the comment into the implementation. Suggested fix: reword the comment to reference the case-sensitive slice(0,7) prefix check that must stay in lockstep with apiKeyAuth.

@gitwire-hq

gitwire-hq Bot commented Nov 20, 2026

Copy link
Copy Markdown
Contributor

✅ GitWire: A review is already published for this head — no new review was posted. Outcome: NEEDS_DISCUSSION.

@Alajmah

Alajmah commented Nov 20, 2026

Copy link
Copy Markdown
Contributor Author

Superseded by #463 (v8): review findings 2 and 4 fixed structurally (shared extractBearerToken + corpus contract test), demo-dashboard manifest synced to the lockfile's 15.5.28, findings 1/3/5/6 adjudicated with evidence in the new PR body. Head chain: b1aca04 → e4b17b7.

@Alajmah Alajmah closed this Nov 20, 2026
Alajmah added a commit that referenced this pull request Nov 20, 2026
…v9 (#464)

* fix(limiter): complete pre-auth family coverage + advisory remediation (next 16.4.6, body-parser, zod)

PRE-AUTH EXTENSION (maintainer finding on #460): /api/auth/check
(session-cookie Redis lookup) and /api/bootstrap/state + /first
(anonymous DB work) mount before apiKeyAuth and carried the same
bucket-rotation mechanism. The pre-auth match now covers the COMPLETE
/api/auth/* and /api/bootstrap/* families (prefix regex, case +
trailing-slash tolerant). Bearer identity unchanged for ordinary
authenticated routes. Three new regressions: AUTH-CHECK end-to-end
(130 rotating-Bearer session requests → 429@121 ONE bucket);
BOOTSTRAP-STATE end-to-end (130 anonymous → 429@121 ONE bucket);
FAMILY VARIANTS (8 path forms all key on the IP).

THREE-LAYER REMEDIATION (carried): session-cookie rotation closed
(apiKeyAuth-exact recognition); pre-auth login closed (IP-keyed);
route-variant closed (trailing slash + case); derivation isolated
(derived as an exported function — the CodeQL missing-rate-limiting
model pairing resolved without behavior change).

ADVISORY REMEDIATION (the #457 master-push CI failed on seven new
advisories, three critical, on next ≤16.4.0):
- next 16.3.8→16.4.6 (dashboard; GHSA-2464 critical, GHSA-673q
  critical, GHSA-p3m5 high): requires middleware.ts→proxy.ts rename
  (function export renamed too) + @next/swc platform binaries
  hand-carried to 16.4.6 (the version mismatch crashed builds with
  'Cannot read properties of undefined (reading project)').
- next 15.5.24→15.5.28 (landing; same advisories' 13.x-15.x range).
- body-parser 1.20.6→1.20.9 (override; GHSA-27r3 high).
- zod 3.25.76→4.6.7 (web; GHSA-mrhx high).
Lockfile entries hand-carried with registry integrities (npm on
Windows refused to re-resolve overrides + platform binaries — the
known quirk).

VERIFICATION: audit gate 0 high/critical; limiter identity (17) +
behavior (9) + webhook body-limit (14) = 40 tests green; dashboard
5 suites / 66 tests green; dashboard build verified locally
(compiled successfully, 29/29 static pages).

Signed-off-by: Alajmah <55954113+Alajmah@users.noreply.github.com>

* fix(limiter): structural bearer-recognition lockstep + corpus contract; demo-dashboard manifest sync

Adjudication corrections for the v7 chain (#462 review):

- apiKeyAuth and deriveLimiterIdentity held unlinked copies of the same
  Bearer recognition rule, kept in sync only by comments. Both now call a
  shared extractBearerToken() (bearerToken.js), and a 12-case corpus
  contract test pins the one rule both mechanisms depend on — divergence
  in either direction becomes a test failure, not a silent re-opening of
  the bucket-rotation bypass.
- Fixed a garbled test comment that cited a nonexistent ^Bearer\s+ regex
  for the tab-scheme case; apiKeyAuth actually uses a literal starts-with
  "Bearer " check, byte-identical to the limiter's recognition.
- demo-dashboard manifest pinned next 15.5.24 while the hand-carried
  lockfile resolves 15.5.28 — any npm install would have silently
  reverted the advisory floor. Manifest now matches the lockfile (npm ci
  dry-run verified); demo dashboard build verified green under 15.5.28.

Limiter semantics unchanged: 38 limiter tests + 15 triage-pr-guards green.

Signed-off-by: Alajmah <55954113+Alajmah@users.noreply.github.com>

* fix(auth): remove stale query-param claim from apiKeyAuth comment

The comment said 'Extract key from header or query param' but no
query-parameter authentication path exists (nor has one in the current
implementation's history) — the header is the only in-band key channel.
Confirmed against the code: zero req.query reads in auth.js. The stale
line drew a review finding ('query-param auth has no limiter identity
counterpart') whose premise was the comment, not the code.

Signed-off-by: Alajmah <55954113+Alajmah@users.noreply.github.com>

---------

Signed-off-by: Alajmah <55954113+Alajmah@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant