CVE-2026-12624: 
HashiCorp Vault 脆弱性の分析と軽減

CVE-2026-12624 is an incorrect authorization vulnerability in HashiCorp Vault's ACL policy engine that allows authenticated tokens to bypass wildcard (glob) deny rules when performing LIST requests with a trailing slash on a denied path. This flaw enables a token holding a broader allow rule alongside a narrower wildcard deny rule to enumerate entry names beneath paths it should be denied access to. It affects Vault Community Edition versions before 2.0.3 and Vault Enterprise versions before 2.0.3, 1.21.8, 1.20.13, and 1.19.19. The vulnerability carries a CVSS v3.1 base score of 4.3 (Medium) (HashiCorp Advisory, ENISA EUVD).

The root cause is classified as CWE-863 (Incorrect Authorization): Vault's ACL policy engine failed to consistently apply wildcard (glob) deny rules when a LIST request was made with a trailing slash appended to the denied path. The trailing slash caused the policy engine to strip or misinterpret the path, bypassing the intended deny rule and allowing the request to proceed under a broader allow rule. Exploitation requires a low-privileged authenticated token with a policy configuration that combines a broader allow rule with a narrower wildcard deny rule — a common pattern in least-privilege Vault policy designs. No public proof-of-concept code has been identified at this time (HashiCorp Advisory, ENISA EUVD).

Successful exploitation allows an authenticated attacker to enumerate the names of secrets or entries beneath paths they were explicitly denied access to, resulting in a low-level confidentiality breach. While the vulnerability does not grant read access to secret values themselves, path enumeration can expose sensitive organizational structure, secret naming conventions, and facilitate targeted follow-on attacks. Integrity and availability are not impacted (HashiCorp Advisory).

No public exploit code or evidence of in-the-wild exploitation has been reported for CVE-2026-12624. The EPSS score is 0.0, indicating very low current probability of exploitation, and it does not appear in the CISA Known Exploited Vulnerabilities (KEV) catalog. Exploitation requires a valid, low-privileged Vault token and a specific policy configuration combining broad allow and narrow wildcard deny rules, limiting the attack surface to environments with such policy patterns (ENISA EUVD, HashiCorp Advisory).

1. Reconnaissance: Obtain a valid Vault token with a policy that includes both a broader allow rule (e.g., path "secret/*" { capabilities = ["list"] }) and a narrower wildcard deny rule (e.g., path "secret/sensitive*" { capabilities = ["deny"] }).
2. Identify target path: Determine the path subject to the wildcard deny rule that the attacker wishes to enumerate (e.g., secret/sensitive/).
3. Craft trailing-slash LIST request: Issue a Vault API LIST request to the denied path with a trailing slash appended (e.g., GET /v1/secret/sensitive/?list=true or LIST /v1/secret/sensitive/) using the token.
4. Bypass deny rule: The ACL policy engine fails to match the trailing-slash path against the wildcard deny rule, instead applying the broader allow rule and returning a list of entry names beneath the denied path.
5. Enumerate secrets: Use the returned entry names to map the secret structure under the denied path, enabling targeted follow-on reconnaissance or privilege escalation attempts (HashiCorp Advisory).

• Logs: Vault audit logs showing LIST requests to paths ending with a trailing slash (e.g., /v1/secret/sensitive/) that return successful responses (http_status_code: 200) for tokens whose policies include a deny rule on that path pattern.
• Network: Repeated LIST API calls to Vault endpoints with trailing slashes from a single token or source IP, particularly targeting paths that should be restricted.
• Behavioral: Tokens performing LIST operations on paths outside their expected access scope, especially where the policy includes wildcard deny rules — detectable by cross-referencing audit log access patterns against configured ACL policies (HashiCorp Advisory).

HashiCorp has released patched versions addressing this vulnerability: Vault Community Edition 2.0.3, and Vault Enterprise 2.0.3, 1.21.8, 1.20.13, and 1.19.19. Organizations should upgrade to one of these fixed versions as the primary remediation. As an interim measure, administrators can review and audit ACL policies that combine broad allow rules with narrower wildcard deny rules, and consider temporarily restricting LIST capabilities on sensitive paths until patching is complete (HashiCorp Advisory).

The vulnerability was disclosed by HashiCorp via their security advisory forum (HCSEC-2026-26) on August 10, 2026. No notable independent researcher commentary or significant social media discussion has been identified at this time, consistent with the vulnerability's medium severity and limited exploitation potential.