FAQ

One closed loop: versioned Stack Releases, a private Golden Set, managed qualification, and a workload-level release decision. Fullbeam runs approved work in isolated, ephemeral managed environments with customer-owned provider credentials. It prepares rollout intent but never deploys the customer setup. Access is invitation-only during private preview.

No. Runtime, CI, review, and delivery signals are included only as evidence for a Stack Release qualification or externally supplied production confirmation. Fullbeam does not sell PR review, generic management dashboards, broad organizational profiles, or a general observability log explorer.

No. Fullbeam reports observed associations with the denominator, sample size, and telemetry coverage visible. It does not turn those comparisons into causal productivity claims or engineer scores.

Generic DORA and SPACE dashboards, productivity scores, public builder archetypes, global benchmarks, prompt routing, leaderboards, a public prompt marketplace, and automatic bug review are not part of focused v1.

GitHub is the source of truth for repositories, commits, pull requests, and review outcomes. Fullbeam has observer paths for Claude Code, Codex, and OpenCode; coverage is displayed per client and workflow. Missing runtime telemetry remains an explicit evidence gap rather than being inferred.

Invited teams can create a workspace and configure a qualification with Fullbeam-assisted admission. Model usage is charged to the team's provider account. Fullbeam plan limits and any assisted onboarding terms are shown before a paid commitment; no seat price is presented as validated.