Research dossier
I study AI agents where software systems, security boundaries, and operational evidence matter. The aim is to make agent behavior inspectable before it is trusted. Current work combines reliability-agent evaluation, security audits for agent choices, and developer tools that make context and software evolution easier to examine.
Publication recordCVScholar profile
Evidence anchors
01 · Reliability-agent evaluation
How can we evaluate an SRE agent in a live software environment without mistaking a shallow recovery for a reliable one?
Through UIUC++ SRSE 2026, I am working on SREGym, a benchmark for evaluating SRE agents in live system environments and real-world reliability problems.
02 · ML ecosystem security
What risks emerge when model-hosting platforms load community code and artifacts at ecosystem scale?
A cross-platform empirical study examines custom model-loading risk across five ML hosting ecosystems, using static analysis, malware-signature scanning, and more than 600 developer discussions.
03 · Agent security
Can a trigger steer an agent toward an attacker’s preferred valid choice while its final answer still appears acceptable?
SHIFT compares matched tasks with and without a known trigger, then asks whether a changed valid choice favors an attacker’s target after accounting for ordinary option quality.
The manuscript is not public; this record is deliberately limited to an abstract-level summary.
04 · Coding-agent systems
How can coding agents preserve exact session facts, inspect the right files, and leave testable repair evidence?
ContextLedger measures what survives context compaction, ctxhelm and HelmBench evaluate retrieval efficiency, and PatchSmith records bounded repair runs with sandbox validation and inspectable artifacts.
05 · Agent memory
Can coding agents preserve architectural intent while developers work through fast natural-language programming loops?
This emerging direction studies how agents should use commit history, prior edits, repository memory, and design intent without losing the speed of vibe coding.
06 · Multi-agent verification
How can specialized agents generate relational schemas while formal checks and targeted repair keep the result consistent?
VeriSchema uses agents for entity extraction, relationship mining, normalization, and repair, with Z3 checks and component-level retries for relational consistency.
Supporting systems
These systems make the agent’s context, choices, and program-level effects more inspectable.
Collaboration
If you are working on software reliability, agent evaluation, security, or developer tools, I would be glad to compare notes or review an artifact.