Show HN: Apgar – Catch agent regressions without an LLM judge
Hey HN, We build financial AI agents, and Apgar is the tool we wrote to see whether a harness change made an agent better or worse in production. Our offline evals can't cover every kind of question or session real users bring, and the LLM judge we run in Langfuse is too expensive to use on more than a small slice of traffic, so some failures get through. Apgar is a small release health checker that sits on top of our Langfuse data. It picks up the release from the production traces, and for the
Read original source
Source record
Recorded metrics
points · 3 votes
2026-10-30 15:00:05.255Z
cumulative since publication
Research analysis
Project record
Repository: oimo-ai/apgar. Source account: Hey HN, We build financial AI agents, and Apgar is the tool we wrote to see whether a harness change made an agent better or worse in production. Our offline evals can't cover every kind of question or session real users bring, and the LLM judge we run in Langfuse is too expensive to use on more than a small slice of traffic, so some failures get through. Apgar is a small release health checker that sits on top of our Langfuse data. It picks up the release from the production traces, and for the
Observed adoption & comparisons
Source votes: 3 votes (Oct 30, 2026 UTC; cumulative since publication); previous same-scope observation: 3 (Oct 30, 2026 UTC). These observations do not establish demand growth. Recent records to compare: “Show HN: Cerea – yet another self-hosted LLM chat front end” (Show HN, Nov 5, 2026); “Show HN: Lantern Lattice – the call graph from your compiler, for coding agents” (Show HN, Nov 5, 2026). Shared themes are navigation hints; they have not been verified as the same need or independent corroboration.
Adoption decision
Evaluate it against the tool or manual workflow it would replace. The stored record supports discovery; it does not yet establish reliability, lower operating cost or commercial adoption.
Unverified requirements
Not verified: license permissions, required hardware, installation steps, maintenance status and a representative task with its failure cases.
Next test
Choose one real task and a current alternative. Record the commit, environment, input, output, elapsed time and failure behavior; retain the result before recommending adoption.