17 tasks. May 17, 2026 closed at 10.8x weighted leverage across 309.0 human-equivalent hours in 1,723 minutes of wall-clock time. Supervisory leverage came in at 228.9x.
That is 7.7 weeks of human-equivalent throughput in 28.7 hours. The ceiling was 96.0x; the floor was 1.0x. 17 of the 17 entries came from a single project.
Task Log
| # | Task | Human Est. | Claude | Sup. | Factor |
|---|---|---|---|---|---|
| 1 | Origin-extract Phase 3 — populate services/[ip-cluster] with [content generation] code, merged backend, /jobs API + structlog observability, [ip-cluster] CLI, and relocated test surface (522 passing… | 80.0h | 50m | 5m | 96.0x |
| 2 | Audit other Claude's outstanding-work report against AVIAN engine codebase; corrected stale claims and re-estimated effort | 8.0h | 11m | 6m | 43.6x |
| 3 | Cloud deployment plan for [ip-cluster]: distilled Phases 5-7 (SQS+Fargate+Bedrock wiring, frontend refactor, deploy+cutover) + Phase 8 hygiene into a single 194-line plan doc with Mermaid flow diagram,… | 3.0h | 5m | 1m | 36.0x |
| 4 | Persistence audit follow-through: all 4 fixes shipped. (1) DeltaReplicationPublisher fail-loud in cloud profile. (2) HIGH-severity in-flight exam persistence — Alembic 007_active_exams + ActiveExamRow +… | 24.0h | 55m | 1m | 26.2x |
| 5 | Refresh [ip] valuations and content counts across 25 avian-planning docs (business, marketing, research, README, CHANGELOG); rebuild [ip]-portfolio valuation framework ([cost]-230M floor); scrub… | 14.0h | 40m | 4m | 21.0x |
| 6 | Origin-extract Phase 4: delete src/avian/origin + dying [engine subsystem] subdirs + origin_router + 100+ scripts; slim OriginConfig; collapse regression guard; ratchet coverage 81→82; recover 7 over-deleted… | 16.0h | 47m | 1m | 20.4x |
| 7 | Docstring audit Phase 3 (DOC_OVERSELLS rewrite): F1 fix in admin_events.py (module + _live_session_payload docstrings) for asymmetric user fallback (user_name->entity_id; user_email->""); audit re-run… | 3.0h | 10m | 1m | 18.0x |
| 8 | Deterministic docstring-vs-code audit for engine: AST-driven scripts/audit_docstrings.py with 12 categories (structural + intent-vs-impl), per-finding likely_truth heuristic (fix doc / fix code / review). 65… | 12.0h | 40m | 2m | 18.0x |
| 9 | Docstring audit Phase 2 (FP bookkeeping): added EXCLUDED_FINDINGS set + AuditReport.add_finding() to scripts/audit_docstrings.py with 28 exact-tuple exclusions (file, line, symbol, category) retiring the 29… | 3.0h | 10m | 1m | 18.0x |
| 10 | CI hardening (fixed silently-dead nightly leak gate in engine nightly.yml — wrong import path; dropped continue-on-error from memray steps; mirrored nightly to [ip-cluster] with 500MB import baseline) + full… | 6.0h | 20m | 1m | 18.0x |
| 11 | Documentation | 8.0h | 35m | 1m | 13.7x |
| 12 | avian-engine: fix domain reload manifold dupe (if_exists policy) + 8 unit tests + endpoint regression test; live-validated by reloading 38 AWS/GCP/Azure cert packages into running engine | 5.0h | 25m | 6m | 12.0x |
| 13 | Deployment | 4.0h | 25m | 3m | 9.6x |
| 14 | Infrastructure | 80.0h | 540m | 12m | 8.9x |
| 15 | [engine subsystem] zero-sweep diagnosis: fixed current_day/[scoring model] DB sync + zombie 'running' reaper + content-density auditor, traced 365-day exam-plateau to 74% of goals lacking recall foundation | 15.0h | 210m | 8m | 4.3x |
| 16 | Docstring audit Phase 1: deterministic 9-step disposition pass for 30 doc-likely findings (3 batches of 10), with verbatim docstring/code citations, call-site enumeration, and per-finding justification.… | 24.0h | 360m | 20m | 4.0x |
| 17 | [engine subsystem] zero-sweep on reclassified cloud cert packages: engine restart, fixed autopilot_service NameError (missing import os), ran sweep, 2 real terminals (AZ-120 crossed 0.5 readiness=0.509 day 44… | 4.0h | 240m | 8m | 1.0x |
Aggregate Statistics
| Metric | Value |
|---|---|
| Total tasks | 17 |
| Total human-equivalent hours | 309.0 |
| Total Claude minutes | 1,723 |
| Total supervisory minutes | 81 |
| Total tokens | 10,907,000 |
| Weighted average leverage factor | 10.8x |
| Weighted average supervisory leverage factor | 228.9x |
| Human-equivalent weeks | 7.7 |
Analysis
The highest factor of the day came in at 96.0x and the lowest at 1.0x, a spread of 96.0 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 309.0 human-equivalent hours, or 26 percent of the day. No single task dominated the total, so the weighted average is representative.
Supervisory time was 81 minutes against 1,723 minutes of execution, a ratio of about 1 to 21. Supervisory leverage of 228.9x 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.