vLLM
Open-source LLM inference engine with scale-out, cache and multimodal input paths.
3 explicit connections
AI security research / entities
Explicitly connected records in the evidence library.
104 published records · Page 4 of 4
14 matching records on this page
Open-source LLM inference engine with scale-out, cache and multimodal input paths.
3 explicit connections
Researcher on tool-result formatter omissions and agent-verifier error propagation.
2 explicit connections
Open-source MCP server that exposes Windows automation capabilities.
1 explicit connection
Wiz CIRT investigated live cloud intrusions in which the team attributed coordinated activity to AI-directed attacker infrastructure. Its October 19 reconstruction documents remote access, cloud identity escalation, workload token abuse and exfiltration paths across AWS, GCP, Kubernetes and CI/CD. The report is investi
2 explicit connections
Open HTTP 402 payment flow used by autonomous software purchasing APIs, with client-side signing and facilitator-side verification.
4 explicit connections
Researcher on the Incomparable Minima in Agent Tool Authorization study.
2 explicit connections
JFrog researcher associated with the 42-repository testmetest investigation.
3 explicit connections
Researcher on KnotScryer multi-skill capability-flow security.
2 explicit connections
Researcher on trigger-neutralized counterfactual scoring for code-model poisoning.
2 explicit connections
Researcher on NTBench near-target prompt-injection attacks.
2 explicit connections
Researcher on PROTEA robot plan-safety analysis and MuTRAP robot-planning backdoors.
3 explicit connections
Researcher on TraceTrial-Bench agent-harness execution correctness.
2 explicit connections
Researcher on CONSTRAINTDIAL smart-contract benchmark contamination.
2 explicit connections
MCP integration for Zotero bibliographic workflows, including the generic add-by-URL metadata path examined in CVE-2026-108583.
2 explicit connections
No matches on this page.