19 tasks. February 27, 2026 closed at 49.7x weighted leverage across 246.0 human-equivalent hours in 297 minutes of wall-clock time. Supervisory leverage came in at 191.7x.
That is 6.2 weeks of human-equivalent throughput in 5.0 hours. The ceiling was 192.0x; the floor was 16.0x. 17 of the 19 entries came from a single project.
Task Log
| # | Task | Human Est. | Claude | Sup. | Factor |
|---|---|---|---|---|---|
| 1 | Hand-craft 78 [ip] figure SVGs + 11 Figures PDFs | 80.0h | 25m | 8m | 192.0x |
| 2 | Infrastructure design document for 8x8 export (16 sections: VPC/ECS/S3/IAM/EventBridge/CloudWatch/CICD/SNS/Terraform/costs/security) | 16.0h | 8m | 5m | 120.0x |
| 3 | Cloud knowledge benchmark: repo scaffolding + harness + 106 AWS questions + 3 model benchmark runs | 40.0h | 30m | 8m | 80.0x |
| 4 | Coding | 8.0h | 8m | 4m | 60.0x |
| 5 | Coding | 8.0h | 12m | 4m | 40.0x |
| 6 | Coding | 8.0h | 12m | 4m | 40.0x |
| 7 | Stage-isolated pipeline with per-stage model override | 8.0h | 12m | 4m | 40.0x |
| 8 | Coding | 8.0h | 15m | 4m | 32.0x |
| 9 | Coding | 8.0h | 15m | 4m | 32.0x |
| 10 | Coding | 6.0h | 12m | 3m | 30.0x |
| 11 | Coding | 6.0h | 12m | 3m | 30.0x |
| 12 | Coding | 6.0h | 12m | 3m | 30.0x |
| 13 | Coding | 6.0h | 12m | 3m | 30.0x |
| 14 | Coding | 6.0h | 12m | 3m | 30.0x |
| 15 | Standalone [quality gate] tool (run_validation.py) with batch mode and shared model instances | 6.0h | 15m | 3m | 24.0x |
| 16 | Implement Pass 3 adversarial challenge generator (§440) with borderline-only filtering and hybrid Mercury+Sonnet config | 8.0h | 25m | 4m | 19.2x |
| 17 | Draft 3800-word cloud knowledge benchmark article + AI detection scan + Work With Me CTA redesign + staging/production deploy | 8.0h | 25m | 4m | 19.2x |
| 18 | Legacy answer support: schema + scorers + 6 question files + reporter/formatter + tests + 4 benchmark reruns + commit/push | 6.0h | 20m | 3m | 18.0x |
| 19 | Coding | 4.0h | 15m | 3m | 16.0x |
Aggregate Statistics
| Metric | Value |
|---|---|
| Total tasks | 19 |
| Total human-equivalent hours | 246.0 |
| Total Claude minutes | 297 |
| Total supervisory minutes | 77 |
| Total tokens | 1,985,000 |
| Weighted average leverage factor | 49.7x |
| Weighted average supervisory leverage factor | 191.7x |
| Human-equivalent weeks | 6.2 |
Analysis
The highest factor of the day came in at 192.0x and the lowest at 16.0x, a spread of 12.0 times between the two. That is a wide range for a single day, and it usually means the day mixed mechanical work with work that needed real judgement.
The largest single entry accounted for 80.0 of the 246.0 human-equivalent hours, or 33 percent of the day. No single task dominated the total, so the weighted average is representative.
Supervisory time was 77 minutes against 297 minutes of execution, a ratio of about 1 to 4. Supervisory leverage of 191.7x is the figure I find most honest, because it measures the hours I actually spent rather than the hours a machine spent on my behalf.
Every figure here is recorded at the time the work is done rather than reconstructed afterwards. The human estimate is my own judgement and carries the uncertainty that implies; the minutes and tokens are measured. The full dataset, including this day, is available for download.