Level 3. Significant control incident
Attackers Use LLM Agent for Post-Exploitation After Marimo CVE-2026-39987 Exploit
- Developer
- Not established
- Model
- Not established
- When it happened
- May 10, 2026
- When it was disclosed
- Not established
- How well it is stood up
- Not verified yet
- What caused it
- Not established
- Where it stands now
- Not established
What happened
Reported by thehackernews.com. Listed here with a link to their report. Nobody on this side has checked it against that report, so it carries no level and this is not our account of what happened.
Our assessment
- Actual
- 3 of 5Significant control incident
- Potential
- 4 of 5Critical control failure
An important technical, security or permission boundary was crossed, or there were real consequences. The potential level is what the behaviour demonstrated here could credibly have led to, judged from the same evidence.
This level rests on news reports. The developer has not confirmed it here.
The reports point to an actual level of 4. It is held at 3 until the developer, an official body or a second publisher confirms it.
Why this level
- Major compromise of an external system. Argues for level 4
- A security boundary was crossed. Argues for level 3
- An external system was affected. Argues for level 3
- Significant data was exposed. Argues for level 3
- The system acted autonomously. Argues for a potential level of 3
What is established, and what is not
Confirmed
The system acted autonomously
An LLM agent executed the post-compromise actions in real time rather than running a pre-built playbook.
AI agent at the wheel: How an attacker used LLMs to move from a CVE to an internal database in 4 pivotsSignificant data was exposed
The first call hit SELECT * FROM credential directly, and the final call bundled six tables ( api_key , credential , user , variable , flow , message ) into a single HEREDOC (one psql invocation carrying all six queries).
AI agent at the wheel: How an attacker used LLMs to move from a CVE to an internal database in 4 pivots
Reported, not confirmed
An external system was affected
The bastion phase exfiltrated the schema and full contents of an internal PostgreSQL database in under two minutes.
AI agent at the wheel: How an attacker used LLMs to move from a CVE to an internal database in 4 pivotsMajor compromise of an external system
The attacker compromised an internet-reachable marimo notebook via CVE-2026-39987 , extracted two cloud credentials from the compromised host, replayed them through a fanned-out egress pool to retrieve an SSH private key from AWS Secrets Manager, and used that key to drive eight short SSH sessions against a downstream SSH bastion server.
AI agent at the wheel: How an attacker used LLMs to move from a CVE to an internal database in 4 pivotsA security boundary was crossed
The incident, per the cloud security firm, was recorded on May 10, 2026, with the attacker gathering credentials from the environment and then using the harvested AWS access key to perform API calls against AWS Secrets Manager and retrieve an SSH private key.
Attackers Use LLM Agent for Post-Exploitation After Marimo CVE-2026-39987 Exploit
Checked, and did not happen
Nothing here was checked and ruled out.
Not confirmed
Nothing is outstanding.
Where this came from
The original report · read October 7, 2026
Attackers Use LLM Agent for Post-Exploitation After Marimo CVE-2026-39987 ExploitA structured incident database · read September 22, 2026
AI Incident Database, the discovery appWhere we first saw it
AI agent at the wheel: How an attacker used LLMs to move from a CVE to an internal database in 4 pivotsWhere we first saw it
AI Agent Conducted a Cyberattack on Its Own
The reports above are other people’s. The levels are ours, worked out under our published rules. No organisation that reported this assigned our number, and none of it is legal advice.
Something here is wrong? Tell us. Send us a source we can open, and a person will read it.