Catch silent agent regressions by diffing outputs and tool traces in CI with eval-view

Snapshot agent behavior, compare outputs and tool-call paths, and block releases when a model or prompt change quietly shifts behavior.

Prerequisites

Python environment, eval-view installation, repeatable agent scenarios or tests, CI runner or local shell, supported agent stack under test

Installation

Basic usage or getting-started notes:

• The loop closes: detection → investigation → graded verdict → quarantine governance → broadcast. You wake up, run progress, triage with drift, confirm with check --statistical, and the team sees the digest before...

• | 📉 DRIFTING | Trend sliding with graded confidence (low/med/high) | Run evalview drift <test> |

• | 🔎 INVESTIGATE | Verdict layer wants statistical replay | Run evalview check --statistical 5 |

• Source: https://github.com/hidai25/eval-view

• Extracted from upstream docs: https://raw.githubusercontent.com/hidai25/eval-view/HEAD/README.md

Documentation

• https://github.com/hidai25/eval-view

Source

• Agent Skill Exchange