AgentLens: Production-Assessed Trajectory Reviews for Coding Agent Evaluation
For developers of coding agents, AgentLens provides a more informative evaluation method that captures nuanced agent behaviors, addressing the limitation of binary pass/fail benchmarks.
AgentLens introduces a benchmark for coding agents that evaluates the full trajectory of agent behavior, not just task completion, using formal verification and LLM-based reviews. It enables detailed diagnosis and regression detection in production pipelines.
We present AgentLens, a production-assessed benchmark for interactive code agents. Most code-agent benchmarks reduce a run to a single bit -- did the task pass? -- but the people who actually use these agents experience the entire trajectory: how the agent follows instructions, uses its tools, verifies its own work, recovers from mistakes, and talks to them along the way. AgentLens evaluates that whole trajectory. It pairs formal verification, where an objective check exists, with LLM-written trajectory reviews and side-by-side comparisons, so that each run yields a readable explanation of why the score is what it is. This makes AgentLens useful for more than ranking models: we use it to diagnose model behavior, compare successive versions of our own agent, and catch product regressions in a nightly evaluation pipeline. We release the benchmark as open source at https://github.com/agent-lens/agent-lens-bench.