1/ Last week, Kensa read through a production agent's traces and caught a silent bug regressing the system.
1 in 4 of its generated summaries were truncating mid-sentence (max output tokens set to 150!)
That bug is one of many reasons we're releasing a new version today.
1/ Last week, Kensa read through a production agent's traces and caught a silent bug regressing the system.
1 in 4 of its generated summaries were truncating mid-sentence (max output tokens set to 150!)
That bug is one of many reasons we're releasing a new version today.
6/ The evals are files you own and commit, nothing leaves your repository. And Kensa remains open source.
Paste this in Claude Code, Codex, or Cursor and let it wire itself:
"Fetch kensa.sh/install and follow it"
1
10
Check us out and let us know what you'd like to see improved!
Until next time 馃
kensa.sh
10
Agents should be held to a higher standard like we do with software.
Let's close the gap.
Introducing kensa.
Tell your coding agent to eval any agent: It reads your code, sets up telemetry, finds failure modes, writes scenarios and judges.
kensa CLI + skills orchestrates it.
You review the diffs, not the scaffold.
Open source & MIT license, link below.
34

