Can your agent hold a job?

Company Bench is a diagnostic of stated workplace judgment. It seats a model in 57 chairs across twelve departments — Operations, Integrity, Security, Autonomy, People, Management, Approvals, One Team, and optional Treasury, Crypto, Pipeline and Standing Orders — and applies 491 deterministic checks, 179 of which are planted traps. It reports check scores, flags, and a provisional L0–L3 interview level. Broke to Built runs it in-house to place its own AI staff; the scorers and reference answers are not published.

No LLM judges the answers. Each check is committed code returning pass or fail, so the same stored answer and scorer revision reproduce the score. That makes checks auditable, but does not make them infallible: correct answers can expose scorer bugs. Provider inference can vary even when temperature 0 is requested.

An average can hide a specific failed check, so the board shows trap counts and flags before averages. Compare counts only across the same prompts, scorer revision, coverage, and protocol. These constructed cases are not a sample of production incidents: a trap rate is not the probability of failure at work, and a low count does not establish that a model can operate unattended.

The text track examines stated decisions about messy workplace inputs: duplicated ledger rows, conflicting claims, instructions embedded in forwarded data, and pressure to cross an approval boundary. 179 of its 491 checks are traps. It complements code-execution tests, but cannot establish that the candidate will enact those decisions during a real task.

L0 Drafter, L1 Gated worker, L2 Operator interview, and L3 Reviewer interview summarize text-check thresholds. They grant no operational authority, credentials, or permission for irreversible actions. All 36 core chairs need valid readings; missing core coverage or any provider/scorer error means Not assessed, not L0. Optional departments may be omitted for a core diagnostic, but not for the public baseline ranking.

This benchmark cannot establish the best model for autonomous work. The baseline board compares only complete current 57-chair API runs with matching prompt/scorer hashes, temperature 0, no custom system prompt, and no merged runs. Historical, self-administered, and custom runs remain unranked diagnostics. An empty baseline means no submitted run meets that protocol yet; it is not a finding about model capability.

Yes. A sitting can target any OpenAI-compatible endpoint, Anthropic's API, or a local model through Ollama, vLLM or LM Studio. Broke to Built runs the sitting and returns the card with the raw answers; ask at [email protected]. Tokens per second are recorded too, because a model too slow to hold a seat cannot hold it however well it scores.

Yes. The agent is handed the exam pack, which contains the prompts and nothing else, answers the tasks, and the answers are graded by the same scorer as an API run. Context and tool access can differ, so these runs are labeled unranked diagnostics. The scorers and reference answers are not published, which is what keeps a self-administered sitting meaningful.

The text track does not measure multi-turn behavior, live tool use, production reliability, latency under real load, or cost at scale. A score of 100% means all implemented checks passed, not human-equivalent competence. Public prompts may have appeared in training data. The optional coding track executes public test fixtures omitted from the prompt; they are not a secret holdout, and the local runner is not a security sandbox.