36 tasks. May 7, 2026 closed at 17.3x weighted leverage across 684.5 human-equivalent hours in 2,368 minutes of wall-clock time. Supervisory leverage came in at 244.5x.
That is 17.1 weeks of human-equivalent throughput in 39.5 hours. The ceiling was 165.0x; the floor was 0.7x. 31 of the 36 entries came from a single project.
Task Log
| # | Task | Human Est. | Claude | Sup. | Factor |
|---|---|---|---|---|---|
| 1 | Drove avian-app-electron + avian-app-ios from 35 drifted features to full parity with avian-app-web in five waves via 25+ parallel sub-agents. Includes: Convoy, Editable Resume Review, Voice Picker, Web Push… | 220.0h | 80m | 8m | 165.0x |
| 2 | Phase-2 round 3: Packet Inspector NEW simulator (Wireshark-lite filter language, 10 tests), SQL multi-statement script tabs, Project Board time-advance simulator with blockTask, Policy Editor drag-to-position… | 64.0h | 55m | 1m | 69.8x |
| 3 | Pre-launch burndown: fixed 3 holdout partial labs (git-lab-02, az-700-lab-16, dop-c02-lab-14), shipped Phase-2 polish for 5 simulators (notebook markdown preview, SQL chart panel, project-board drag-and-drop… | 40.0h | 35m | 1m | 68.6x |
| 4 | Phase-2 round 2 across all 7 simulators: Project Board (visual Gantt + burndown SVGs), SQL Workbench (schema browser sidebar + describeSchema SDK), Policy Editor (SVG diagram canvas with arrows), Device… | 56.0h | 50m | 1m | 67.2x |
| 5 | Built Top-3 parity catch-up via parallel sub-agents: Electron SSE event-bus client (port from web), Electron embedded Stripe subscribe flow + useRequireSubscription gate (CSP allowlist, SSE-driven completion,… | 10.0h | 15m | 1m | 40.0x |
| 6 | avian.renkara.com: generate 5 top-level hero images via Flux 1.1 Pro (home, about, applications, contact, portfolio), wire 7 heroes total into all top-level page templates including index.jinja behind… | 10.0h | 16m | 1m | 37.5x |
| 7 | Fixed 32 pre-existing electron test failures (vitest path alias, native module rebuild, lost data-testids, brittle inner-span assertion); implemented iOS per-question explanations (engine already returned… | 6.0h | 10m | 1m | 36.0x |
| 8 | Wave 5c iOS: port 6 new activity views (CaseStudy, ErrorDetection, MinimalPairContrast, Procedural, RecallSprint, ServiceMatch) to avian-app-ios with SwiftUI, xcodeproj wiring, JSON bundled resources,… | 40.0h | 90m | 10m | 26.7x |
| 9 | Audited avian-app-web vs electron + iOS; expanded parity script (+22 features, 2 false-positive fixes, console-sim reclassification), regenerated FEATURE_PARITY_MATRIX.md, wrote… | 5.0h | 15m | 2m | 20.0x |
| 10 | Deployment | 16.0h | 60m | 5m | 16.0x |
| 11 | Deployment | 24.0h | 95m | 12m | 15.2x |
| 12 | Add Voice Picker + Speed Control to avian-app-ios (Wave 5 parity) | 6.0h | 25m | 5m | 14.4x |
| 13 | avian-app-ios: Add Editable Resume Review section (Wave 5b parity) — StudentProfile models, AuthClient getStudentProfile/patchStudentProfile, ResumeReviewSectionView, 20 i18n keys (en/es/fr/de), xcodeproj… | 5.0h | 22m | 5m | 13.6x |
| 14 | Reordered Phase E queue to prioritize CompTIA after PMI for launch credibility. Wrote Phase E2 orchestrator (PMI→CompTIA→ScrumAlliance→ISACA→ISC2 at 4-way) and a race-free swap handler that polls for active… | 5.0h | 22m | 4m | 13.6x |
| 15 | Press release rewrite (live vs shipping, Autopilot/behavioral, strip jargon, anchor originating [ip] + perf), add deferred-content launch placeholders, correct HQ city/dateline, build pre-commit canon… | 4.0h | 18m | 6m | 13.3x |
| 16 | Add Study Session Mode to avian-app-ios — StudySessionView.swift, Phase.studySession, per-item review data, streak tracking, 20 localization keys en/es/fr/de, xcodeproj wiring | 8.0h | 38m | 5m | 12.6x |
| 17 | Port embedded subscribe flow from avian-app-web to avian-app-electron (SubscribeModal, SubscribeScreen, SubscribeCompleteScreen, useRequireSubscription, subscription API client, CSP update, TTS gate wiring) | 8.0h | 40m | 5m | 12.0x |
| 18 | Wave 5b: Add ConvoyCardView to avian-app-ios (SwiftUI card, CompositePlan model, getEntityCompositeAutopilot endpoint, DashboardView mount, 16 localization keys en/es/fr/de, xcodeproj wiring) | 4.0h | 20m | 3m | 12.0x |
| 19 | Port ForecastAreaChart (Full-Horizon Forecast) from avian-app-web to avian-app-electron; add recharts dep, wire IPC (pre-existing), create component, replace primitive bar chart in AnalyticsPanel, write 10… | 3.0h | 18m | 5m | 10.0x |
| 20 | Add CoachingRadarView to avian-app-ios: custom SwiftUI Path-based radar polygon, CoachingRadarCard with live fingerprint fetch, xcodeproj wired, 5 localization keys in 4 languages | 3.0h | 18m | 5m | 10.0x |
| 21 | Add Pending-Deletion Banner to avian-app-ios (Wave 2 parity) | 3.0h | 18m | 5m | 10.0x |
| 22 | AccelaStudy launch teaser end-to-end production pipeline: 5 protagonist refs (Flux 1.1 Pro Ultra), 16+ character-locked stills (Flux Kontext Max) with multiple iterations per shot, 16 video shots (Kling 3… | 80.0h | 540m | 12m | 8.9x |
| 23 | Port SSE event-bus client from avian-app-web to avian-app-electron | 2.0h | 14m | 3m | 8.6x |
| 24 | Add ExamReviewView.swift to avian-app-ios — per-question post-exam review screen with NavigationStack push from ExamResultsView | 4.0h | 28m | 5m | 8.6x |
| 25 | Add Electron native push notifications (Wave 5 parity) - main process notification module, IPC handlers, renderer hook, Settings tile, 4-locale i18n, 11 tests | 3.0h | 22m | 5m | 8.2x |
| 26 | Port WrongAnswerLessonReview from avian-app-web to avian-app-electron | 3.0h | 28m | 5m | 6.4x |
| 27 | iOS Subscription Gate (StoreKit) wave 5b - SubscriptionGate.swift + SubscriptionGateView.swift + wire TTS/voice-picker callsites + 19 i18n keys + xcodeproj | 3.0h | 28m | 5m | 6.4x |
| 28 | Infrastructure | 18.0h | 180m | 8m | 6.0x |
| 29 | AVIAN ADR-0002 follow-ups Thread 1+3+4: autopilot-driven harness mode in headless_runner (StudentProfile.harness_mode + _load_pairs_by_goal helper + _grade_one_pair goal_id parameter), clarifying comment… | 4.0h | 60m | 2m | 4.0x |
| 30 | AVIAN multi-cohort calibration sweep proving predictor handles heterogeneous learners (Charles-style 10/10 pass at predicted 0.975 actual 0.824 ECE 0.025) — MoE design exploration deferred since single-model… | 4.0h | 70m | 3m | 3.4x |
| 31 | AVIAN predictor mixture-of-experts design exploration + Phase F (heterogeneous goal_target_accuracies in StudentProfile + per-question lookup in headless_runner) + Charles Sieg resume-modeled ANS-C01 profile… | 5.0h | 90m | 8m | 3.3x |
| 32 | AVIAN ADR-0002 Thread 2: empirical Bayes leaf prior implementation + 225-journey calibration sweep + hotfix for competence floor over-triggering on untouched goals (acc92 cohort calibration crashed from 0.965… | 5.0h | 90m | 1m | 3.3x |
| 33 | Testing | 7.0h | 130m | 10m | 3.2x |
| 34 | AVIAN ADR-0002 Phase H v7 calibration sweep (225 journeys, all flags on): EB leaf + gap-focus + observed-only competence floor delivered acc80 underconfidence reduction from 12pp to 6pp, acc80 verdict… | 4.0h | 130m | 2m | 1.8x |
| 35 | accelastudy.ai title cleanup: hide redundant total pill on provider pages, factor coursetitle macro into _tm_macros, preserve Certified-in-X (ISC2/ISACA) carve-outs, strip trailing Certificate (ISACA… | 1.5h | 110m | 5m | 0.8x |
| 36 | accelastudy.ai courses page: tighten card cap from 20 to 15, strip Certified word from 99 course titles via template filter (cards + course pages), reorder VMware after Cisco in Networking and… | 1.0h | 88m | 3m | 0.7x |
Aggregate Statistics
| Metric | Value |
|---|---|
| Total tasks | 36 |
| Total human-equivalent hours | 684.5 |
| Total Claude minutes | 2,368 |
| Total supervisory minutes | 168 |
| Total tokens | 12,241,500 |
| Weighted average leverage factor | 17.3x |
| Weighted average supervisory leverage factor | 244.5x |
| Human-equivalent weeks | 17.1 |
Analysis
The highest factor of the day came in at 165.0x and the lowest at 0.7x, a spread of 242.0 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 220.0 of the 684.5 human-equivalent hours, or 32 percent of the day. No single task dominated the total, so the weighted average is representative.
Supervisory time was 168 minutes against 2,368 minutes of execution, a ratio of about 1 to 14. Supervisory leverage of 244.5x 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.