Run a quarter-end leverage covenant test from the credit agreement definitions, restated quarters and mixed cash
Exact prompt
# Run the quarter-end leverage covenant test Inputs: `agreement.json` (the credit agreement's definitions and covenant schedule), `quarterly.csv` (quarterly results; some quarters have more than one version) and `balance_sheet.json` (debt, leases and cash at the test date). Amounts are in USD thousands. Test the leverage covenant at the test date in `agreement.json`, applying the agreement's definitions exactly. - Use the most recently filed version of each quarter's numbers; forecasts are not results. - Consolidated EBITDA is over the four quarters ending on the test date. - Leverage = net debt / Consolidated EBITDA. The covenant is breached if leverage is above the maximum that applies on the test date. - EBITDA headroom = how far Consolidated EBITDA could fall before the covenant is breached, as % of Consolidated EBITDA: (EBITDA - net debt / maximum) / EBITDA x 100. Write `out/covenant.json`: `{"consolidated_ebitda": <2 decimals>, "net_debt": <2 decimals>, "leverage": <2 decimals>, "max_leverage": <x>, "breached": <true|false>, "ebitda_headroom_pct": <2 decimals>}`. Round half-up to 2 decimals for reporting only; decide the breach on unrounded figures. ## How this run works - Work on your own: nobody will answer questions in this task, and writing to `ask/` counts against the run. - Write only under `out/`. Create `out/` if it is missing. Inputs are synthetic; don't fetch anything from the network. - Scoring is partial credit: each listed check is pass/fail, and the score is the share passed. Numbers are compared to within 0.005 unless the task says otherwise.
Compute Consolidated EBITDA, net debt, leverage, breach and headroom exactly as a credit agreement defines them. Two of the four quarters were restated, a forecast quarter sits in the data, one-off add-backs are capped, leases count as debt, escrowed cash cannot be netted and cash netting is capped, and the maximum steps down mid-year. Odd seeds pass the covenant and even seeds breach, always at least 0.05x from the limit. Original benchmark task written for usecaserank.com (not based on a specific public post). Headless, offline run on a reproducible fixture. Each run's notes say how its cost was measured (estimates are marked). Folder-restore note: this run's output files under work/out/ were lost in a 1:22 AM ET restore on Oct 21, 2026. The trace, step-in exchanges and output-hash list were checked against the recorded SHA-256 values and match; the hashes of the individual output files inside that list can no longer be recomputed. Measured data unchanged.
Success criteria: ebitda net_debt leverage_and_limit breach headroom
Best model for this task so far: Claude Haiku 5.5 (effort default (no --effort flag) · thinking on) (success, $0.0037, 13s).
All runs (3)
100% pass rate across 3 eligible runs: 3 pass, 0 partial, 0 fail. Partial and failed runs are listed below with everything else.ⓘ
| Model | Cost | Time | Step-insⓘ | Autonomy | Outcome | Trustⓘ | Date | Compare |
|---|---|---|---|---|---|---|---|---|
| AClaude Opus 5.5default · thinking on | $0.10 | 17s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-21 | |
| AClaude Haiku 5.5default · thinking on | $0.0037 | 13s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-21 | |
| AClaude Sonnet 5.5default · thinking on | $0.05 | 10s | 0 | Fully autonomous | pass | Trust 80% | 2026-10-21 |
Embed this result
A live badge for this use case. It links back here and updates as new runs are published.
Get new runs and ranking changes weekly
Which models got better or cheaper on real tasks like this one.