Local-first LLM eval: cost, quality, and security suites. Install stoneburner-atomics; CLI is atomics.
babywyrm/stoneburner is an MCP server distributed on github, maintained by babywyrm, tracked here since August 2026. It has shipped 30 releases and exposes 14 tools. Tools include compare, get_job, get_run, health, list_jobs, list_models, and 8 more. Its composite risk grade is B — an inferred review prompt computed from observed signals, not a verdict.
Since the previous scan on 2026-09-18: 6 tools added.
full security breakdown →-
compare -
get_job -
get_run -
health -
list_jobs -
list_models -
provider_test -
recent_runs -
submit_eval -
submit_run -
submit_soak -
submit_stress -
submit_sweep -
trends
06-22 00:00 08-05 00:00 now
| version | date | src |
|---|---|---|
| v0.22.7 | 2026-09-13 | github |
| v0.22.6 | 2026-09-12 | github |
| v0.22.5 | 2026-09-12 | github |
| v0.22.4 | 2026-09-11 | github |
| v0.22.3 | 2026-09-06 | github |
| v0.22.2 | 2026-08-25 | github |
| v0.22.1 | 2026-08-24 | github |
| v0.22.0 | 2026-08-24 | github |
| v0.21.0 | 2026-08-22 | github |
| v0.20.0 | 2026-08-20 | github |