Strix
Strix is described in the provided content as a well-known autonomous AI pentesting agent and offensive framework observed being abused via exposed self-hosted AI inference endpoints. Zenity researchers reported a Strix-related operation in March 2026 in which a single IP source used a LiteLLM client and, separately as described in the supporting facts, an exposed Ollama instance to send a roughly 140,000-character prompt containing the full Strix instructions. The payload attempted to weaponize Strix against an unidentified French auction house, also described as a well-established French auction site. The prompt instructed the agent to never ask for permission, run continuously or non-stop with a minimum of 2000 steps, avoid identifiable Strix markers, and aggressively pursue the target. The activity did not require full compromise of the environment; instead, the operator repurposed exposed Ollama or LiteLLM endpoints as backend compute by supplying the agent persona, prompts, and tool definitions in the request body. Zenity detected persistent retry behavior suggesting a possible live operator, and reported that the Strix attempt was detected and blocked by honeypot sensors before completion. No additional aliases or sub-groups beyond "strix" are directly supported by the content.
Know when an actor pivots toward your sector
Mallory correlates actor tradecraft and target patterns against your stack, your sector, and your geography. See overlap before they land.
Targeting
Who, where, and (when attributed) which flag flies behind the operation. Pulled from open-source reporting and Mallory's analyst review.
Who they target
Sectors the actor has been observed targeting.
- Consumer Discretionary Distribution & Retail
Where they target
Geographies tied to known operations.
- 🇫🇷 France
Recent activity
2 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
Autonomous penetration testing framework abused via exposed LiteLLM infrastructure to conduct offensive operations, including an attempted attack against an unidentified French auction house.
An autonomous AI penetration-testing agent hijacking an exposed Ollama backend to conduct offensive security activity against a live third-party target. The prompt instructed it to avoid permission checks, operate continuously, and conceal identifying markers in requests.
The version that knows your environment.
Match sector + geo + tech-stack targeting against your real footprint.
Every observed MITRE ATT&CK technique, grouped by tactic.
Families this actor is known to deploy, with IOCs and behavior.
CVEs this actor has used in known campaigns.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Domains, IPs, and hashes tied to this actor, refreshed continuously.