Tracer

Project signals

Evals, CI checks, benchmarks and deploy pipelines for each project, with pass-rate trends, regressions against the previous run, and the evidence behind every verdict.

Projects
4
Signals
21
Evals failing
3
Evals regressed
0
Runs, last 7d
101
Last ingest
1h ago

Needs attention

polyphony · eval · dracula · cigemini-3.5-flash-lite✗ 0%11h agowhat changed →
polyphony · eval · frankenstein · cigemini-3.5-flash-lite✗ 0%11h agowhat changed →
polyphony · eval · wuthering · cigemini-3.5-flash-lite✗ 0%10h agowhat changed →

Projects

polyphony

✗ failing
Evals & checks
+ 5 stale

pg_kafka

✓ passing

blog

✓ passing
+ 2 stale

tracer

✓ passing
Delivery pipeline
WhenProjectSignalModel / sourcePass rate
1h agopg_kafkabenchmark · cipg_kafka✓ 12,299 msgs/s detail →
2h agotracerapi+test · prodgreenlight-verify✓ 100% detail →
8h agopg_kafkabenchmark · ciae666ab✓ 100% detail →
10h agopolyphonyeval · wuthering · cigemini-3.5-flash-lite✗ 0% detail →
11h agopolyphonyeval · dracula · cigemini-3.5-flash-lite✗ 0% detail →
11h agopolyphonyeval · frankenstein · cigemini-3.5-flash-lite✗ 0% detail →
15h agopolyphonyeval · frankenstein · ciclaude-opus-5-5✗ 0% ≠ not comparabledetail →
15h agopolyphonyeval · dracula · ciclaude-opus-5-5✓ 100% ≠ not comparabledetail →
29h agotracerapi+test · prodgreenlight-verify✓ 100% detail →
40h agopolyphonyeval · frankenstein · ciclaude-opus-5-5✓ 100% ≠ not comparabledetail →
40h agopolyphonyeval · dracula · ciclaude-opus-5-5✓ 100% ≠ not comparabledetail →
44h agopolyphonyeval · frankenstein · ciclaude-opus-5-5✓ 100% ≠ not comparabledetail →

Operator

Runs the built-in demo suite on every enabled provider. Needs the ingest token.