CVE-2026-56696: OpenHarness remote project-context commands allow persistent prompt poisoning
OpenHarness /issue and /prcomments slash commands lack remoteinvocable=False protection, allowing remote channel senders to write attacker-controlled Markdown into project context files. Admitted remote attackers can inject malicious content into .openharness/issue.md and .openharness/prcomments.md files, which are subsequently injected into runtime system prompts, persistently influencing local agent behavior.
Published Jun 23, 2026 · Updated Sep 4, 2026
Affected packages
openharness-ai
- Ecosystem
- PyPI
- Affected
- <= 0.1.9
- Fixed in
- No patched version yet
- Check
pip show openharness-ai
Advisory details
From the GitHub Advisory Database (CC BY 4.0).
OpenHarness /issue and /pr_comments slash commands lack remote_invocable=False protection, allowing remote channel senders to write attacker-controlled Markdown into project context files. Admitted remote attackers can inject malicious content into .openharness/issue.md and .openharness/pr_comments.md files, which are subsequently injected into runtime system prompts, persistently influencing local agent behavior.
References
- https://nvd.nist.gov/vuln/detail/CVE-2026-56696
- https://github.com/HKUDS/OpenHarness/pull/272
- https://github.com/HKUDS/OpenHarness/commit/27bb93b810e9ea8fa4832eab7152eeb3b4a6bffb
- https://www.vulncheck.com/advisories/openharness-prompt-injection-via-issue-and-pr-comments-slash-commands
- https://github.com/advisories/GHSA-24cw-228q-3rx9
How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology
