Octo STS (github.com/octo-sts/app) before 0.11.2 returns trust policy configuration to callers whose OIDC token does not satisfy that…
CVE-2026-105770
Octo STS (github.com/octo-sts/app) before 0.11.2 returns trust policy configuration to callers whose OIDC token does not satisfy that policy. TrustPolicy.CheckToken included the policy's issuer, subject and audience values or patterns and claimpattern names and regular expressions in its PermissionDenied errors, which Exchange returned verbatim. Exchange also selected the GitHub App installation before checking the token, so errors from resolving the policy's app or apppattern pin were returned to callers the policy does not trust. An attacker who holds any validly signed OIDC token, such as a GitHub Actions token from a repository they own, and who knows or guesses a target owner/repo and identity name can read that policy's matching rules, even when the policy is stored in a private repository. These rules often name internal repositories, branches, environments or cloud account identifiers. The issue does not allow minting tokens or bypassing a trust policy. This issue is fixed in 0.11.2.
Assessment
- CVSS
- 5.3 (v3.1)
- Severity
- Medium
- Known exploited
- Not listed in CISA KEV
- Exploit availability
- Nothing found in Exploit-DB, Metasploit, Nuclei or public PoC repositories
- Published
- 3 November 2026
- Last updated
- 3 November 2026
- Assigner
- chainguard
CVSS vectorCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N
At the source
References
Community proofs of concept
No community proof of concept yet. If you have one for a published vulnerability and the right to share it, add it below — it will appear here once reviewed.
Submit a proof of concept
Only for publicly disclosed vulnerabilities, and only material from testing you were authorised to perform. This platform does not buy, sell, or broker exploits — see earning from vulnerability research for legitimate bounty programs. Submissions are reviewed before they appear. Text only; do not include live credentials, personal data, or anything targeting systems you are not permitted to test.
Query this programmatically
Agents can read this through the MCP endpoint, or over HTTP:
curl https://cve.zerocrew.in/api/cve/CVE-2026-105770