TR-17 (Rev. 05-15-2026)File No. EH-001-2026

Archive-style profile

HAN SEUNG HO

Technical Profile Record

Archive Report Visual Concept

Personal Profile Record

Record Date: 2026

Agentic coding experiment

2026 Productivity Field Log

File EH-WRK-25/26Reviewed

Synopsis

The 2026 objective is to explore how far software execution can be compressed and expanded by placing AI agents at the center of daily work. This field log tracks the resulting work cadence, machine workload, and applied workstreams.

Handling notice: Private repository names, raw commit messages, internal URLs, and implementation details are intentionally omitted. Entries are editorial summaries of recorded activity, not a live GitHub feed.

Experiment telemetry / 2026

Agentic Coding at Operating Scale

Observation A / GitHub execution cadence

The experiment, one working day at a time.

01—194

Recorded contribution events7,214

192 active days / 194 observed

Public commits

106commit contributions

Restricted

7,047private contribution events

Peak day

155Mar 15

Longest run

145dconsecutive active days
Daily GitHub contribution counts
DateContributions
2026-01-010
2026-01-026
2026-01-039
2026-01-042
2026-01-054
2026-01-061
2026-01-072
2026-01-084
2026-01-093
2026-01-104
2026-01-112
2026-01-122
2026-01-133
2026-01-1412
2026-01-157
2026-01-1611
2026-01-178
2026-01-1813
2026-01-198
2026-01-2015
2026-01-2122
2026-01-2244
2026-01-2335
2026-01-2432
2026-01-2579
2026-01-2655
2026-01-2735
2026-01-2848
2026-01-2914
2026-01-3026
2026-01-3112
2026-02-0132
2026-02-0266
2026-02-0323
2026-02-0448
2026-02-0534
2026-02-0638
2026-02-0718
2026-02-0819
2026-02-0910
2026-02-1063
2026-02-1131
2026-02-1217
2026-02-1313
2026-02-1431
2026-02-1514
2026-02-167
2026-02-172
2026-02-180
2026-02-198
2026-02-2053
2026-02-2191
2026-02-2249
2026-02-2377
2026-02-2465
2026-02-2531
2026-02-26130
2026-02-2747
2026-02-2872
2026-03-0164
2026-03-0291
2026-03-0361
2026-03-0425
2026-03-0538
2026-03-0631
2026-03-0724
2026-03-0877
2026-03-0973
2026-03-1040
2026-03-1121
2026-03-1257
2026-03-13134
2026-03-1480
2026-03-15155
2026-03-1660
2026-03-17106
2026-03-1814
2026-03-1939
2026-03-2034
2026-03-2113
2026-03-222
2026-03-2326
2026-03-2438
2026-03-2520
2026-03-2639
2026-03-2732
2026-03-2896
2026-03-2990
2026-03-3075
2026-03-3148
2026-04-0136
2026-04-0260
2026-04-0351
2026-04-0492
2026-04-0573
2026-04-0654
2026-04-0726
2026-04-0816
2026-04-0911
2026-04-108
2026-04-1123
2026-04-1227
2026-04-1310
2026-04-1410
2026-04-156
2026-04-1612
2026-04-1721
2026-04-1812
2026-04-1969
2026-04-2014
2026-04-216
2026-04-225
2026-04-2325
2026-04-2433
2026-04-2512
2026-04-2641
2026-04-2734
2026-04-285
2026-04-298
2026-04-3024
2026-05-0124
2026-05-0271
2026-05-0366
2026-05-0469
2026-05-0566
2026-05-0626
2026-05-0743
2026-05-085
2026-05-0932
2026-05-1028
2026-05-119
2026-05-1245
2026-05-1387
2026-05-1467
2026-05-1551
2026-05-1646
2026-05-1715
2026-05-1817
2026-05-1918
2026-05-2035
2026-05-2145
2026-05-2233
2026-05-2357
2026-05-2458
2026-05-2546
2026-05-2638
2026-05-2713
2026-05-2835
2026-05-2933
2026-05-3046
2026-05-3115
2026-06-0132
2026-06-0229
2026-06-0336
2026-06-0430
2026-06-0544
2026-06-0638
2026-06-0735
2026-06-0832
2026-06-0929
2026-06-1054
2026-06-1144
2026-06-1230
2026-06-1387
2026-06-1428
2026-06-1532
2026-06-1641
2026-06-1724
2026-06-1856
2026-06-1956
2026-06-20117
2026-06-2157
2026-06-2231
2026-06-2329
2026-06-2426
2026-06-2533
2026-06-2631
2026-06-2718
2026-06-2838
2026-06-2940
2026-06-3037
2026-07-0138
2026-07-0252
2026-07-0346
2026-07-0449
2026-07-0571
2026-07-0640
2026-07-0750
2026-07-0837
2026-07-0931
2026-07-1095
2026-07-1149
2026-07-1262
2026-07-137
LessMore

Observation B / agent workload telemetry

How much machine work the workflow puts into motion.

TOK—2026

Total agent tokens processed107.9B

25.8B Codex + 82.1B Claude

Codex25.8B

23.9% share

Claude Code82.1B

76.1% share

CodexClaude Code

Daily volume / square-root scale

Daily agent token volume by tool
DateTotal tokensCodex tokensClaude Code tokens
2026-01-0150418661050418661
2026-01-021008373210100837321
2026-01-031121167720112116772
2026-01-041008373210100837321
2026-01-053025119640302511964
2026-01-0650418661050418661
2026-01-0750418661050418661
2026-01-083529306250352930625
2026-01-091512559820151255982
2026-01-1050418661050418661
2026-01-111008373210100837321
2026-01-1250418661050418661
2026-01-131512559820151255982
2026-01-143529306250352930625
2026-01-151854332020185433202
2026-01-163765061290376506129
2026-01-171575734430157573443
2026-01-184737369310473736931
2026-01-192252501470225250147
2026-01-204282802610428280261
2026-01-213835013250383501325
2026-01-227500834740750083474
2026-01-237952012770795201277
2026-01-246147300650614730065
2026-01-25123509985601235099856
2026-01-26125765875801257658758
2026-01-279305546860930554686
2026-01-288177601790817760179
2026-01-2926130141724432951236868466
2026-01-30595943937116567281479376656
2026-01-3124844107062330133186110937
2026-02-0151111200523111225488000780
2026-02-0254714707521965283525181792
2026-02-0337156285997352897274209962
2026-02-044739182720473918272
2026-02-05549956457122374821427581636
2026-02-0654488493831601469513283469
2026-02-0727393909248529208225409884
2026-02-08311835341128254095183581246
2026-02-092416765770241676577
2026-02-105600389900560038990
2026-02-1144342232439218449404203875
2026-02-1224779301150268886197524125
2026-02-133609027007683088353219612
2026-02-144159625700415962570
2026-02-151510478600151047860
2026-02-1685981090085981090
2026-02-1744152452044152452
2026-02-1818590506018590506
2026-02-1974362024074362024
2026-02-204880007800488000780
2026-02-216971439710697143971
2026-02-229574110540957411054
2026-02-235321532310532153231
2026-02-246041914420604191442
2026-02-256913485145823609685524905
2026-02-2673724559072635004664610586
2026-02-2749016103578846092411314943
2026-02-2885753770944482332813055377
2026-03-0168555762849033932636523696
2026-03-0288794828160205532827742749
2026-03-0370082157127625728673195843
2026-03-0432615592758973141267182786
2026-03-05549526552347829743201696809
2026-03-06100895413393560983973344294
2026-03-07451033630317442237133591393
2026-03-08758801871337072179421729692
2026-03-0993213806983783826294299807
2026-03-1065302491156134454391680368
2026-03-1128008863723293873447149903
2026-03-12850205459360685134489520325
2026-03-1371533861039680719675657891
2026-03-1451015498962230907447924082
2026-03-15747633472302328829445304643
2026-03-16747007460375047111371960349
2026-03-17982310331306495049675815282
2026-03-182306534492762249227891200
2026-03-19363562369114715657248846712
2026-03-20516746562226460998290285564
2026-03-21317601780165674313151927467
2026-03-2216467555110966733055008221
2026-03-23348744027123472266225271761
2026-03-24306681190364209306316981
2026-03-2521564183719183906196457931
2026-03-26592378258385026591993232
2026-03-274413365746509687434826887
2026-03-286576672075426876652240331
2026-03-297597425665344111754398455
2026-03-30824186845331732298492454547
2026-03-31926060377616966566309093811
2026-04-01951868518606563910345304608
2026-04-02811691445427540069384151376
2026-04-03968362806605792968362569838
2026-04-04852803235114714636738088599
2026-04-0590789955910107579897791980
2026-04-06109031670726071931087709514
2026-04-0761893699110337620608599371
2026-04-08646769694366209700280559994
2026-04-091593447061550283984316308
2026-04-1022883915677768390151070766
2026-04-1132208073015622891306457839
2026-04-1252663086251837026474793836
2026-04-132925330543340445289192609
2026-04-1451854373065331432453212298
2026-04-1534901435329607591319406762
2026-04-1651374114221941425491799717
2026-04-1789819655569465496828731059
2026-04-182460295330246029533
2026-04-196042830640604283064
2026-04-2058838574096326674492059066
2026-04-21510781859161160944349620915
2026-04-22407821281278332053129489228
2026-04-23419283050250947054168335996
2026-04-24494361823110210447384151376
2026-04-2517484876066941070107907690
2026-04-26463337440359746058103591382
2026-04-27543241089120242945422998144
2026-04-28520883334361179953159703381
2026-04-2924133769319817461743163076
2026-04-30392992361116748675276243686
2026-05-01365179346212369526152809820
2026-05-02661463269152097204509366065
2026-05-0336605073237792601328258131
2026-05-04633682975231849746401833229
2026-05-05717962783112383128605579655
2026-05-06538272190170736762367535428
2026-05-0748837990297865919390513983
2026-05-0822760326712537595215065672
2026-05-0937098395814427712356556246
2026-05-10589882388199368405390513983
2026-05-112603426560260342656
2026-05-1249534118331252101464089082
2026-05-13864137599100088501764049098
2026-05-14858032917190197409667835508
2026-05-15801429331212828545588600786
2026-05-16489853665269128370220725295
2026-05-1737601801981717626294300393
2026-05-1834097237331833388122638492
2026-05-19374516245148131327226384918
2026-05-2056987113666164694503706442
2026-05-21471481566160202304311279262
2026-05-22371668386213198943158469443
2026-05-23698380727505953547192427180
2026-05-24944177943615919812328258131
2026-05-25569222049240963918328258131
2026-05-26488466229154548475333917754
2026-05-27415981199116021183299960016
2026-05-28534724372138550766396173606
2026-05-2970957468630419932679154754
2026-05-30701865529113264743588600786
2026-05-311641290650164129065
2026-06-01409539435127696908281842527
2026-06-0249396677685018796408947980
2026-06-0329360920661503596232105610
2026-06-0424362659050205248193421342
2026-06-05676428217245374941431053276
2026-06-06475698029177276530298421499
2026-06-073757900360375790036
2026-06-0846605359162631935403421656
2026-06-0951091400468808079442105925
2026-06-1060006349083597671516465819
2026-06-1135311703322040243107679
2026-06-1231466642661254076253412350
2026-06-131345645383952192651250426118
2026-06-14679452756331019257348433499
2026-06-15362017686183199304178818382
2026-06-1671058951727848545682740972
2026-06-1764257424775613837566960410
2026-06-181144080316908982251053182091
2026-06-19815288909258877502556411407
2026-06-2016576088204251337791232475041
2026-06-2162105636358108802562947561
2026-06-22525452068133449979392002089
2026-06-2344209301662781894379311122
2026-06-2438879321032327730356465480
2026-06-25811785096251181944560603152
2026-06-2633646712161770173274696948
2026-06-2772780307131943461695859610
2026-06-28722501276221849499500651777
2026-06-29633477384130363519503113865
2026-06-3066217069118923052643247639
2026-07-0189184025415890265875949989
2026-07-021200086031204480031179638028
2026-07-031040966348168030824872935524
2026-07-041113519272308287311082690541
2026-07-051109538396353961991074142197
2026-07-06570593726172336910398256816
2026-07-0756777370895166891472606817
2026-07-08876722602413244657463477945
2026-07-09878712049447736540430975509
2026-07-10217965124010445277641135123476
2026-07-1117921395371052769347739370190
2026-07-1219617852368195870891142198147
2026-07-13454654936353682058100972878
Daily Codex and Claude Code token volume for the 2026 agentic coding experiment.

Observation C / API-equivalent workload

The operating leverage of the same agent workload.

USD—YTD

Agent workload equivalent$106.3K

$36.5K OpenAI + $69.8K Anthropic / USD

OpenAI / Codex$36.5K

34.3% of total

Anthropic / Claude$69.8K

65.7% of total

OpenAI / Codex$36.5K34.3% / provider API-equivalentAnthropic / Claude$69.8K65.7% / provider API-equivalent
OpenAI / Codex$36.5K

34.3% / provider API-equivalent

Anthropic / Claude$69.8K

65.7% / provider API-equivalent

Model valuation

Claude mix: 96% time-current Opus / 4% Sonnet 4
  1. 01
    Gpt 5.5

    OpenAI / 11.4B tokens / 95.4% cached

    $22.1K20.8%
  2. 02
    Claude Opus 4.6

    Anthropic / 26.8B tokens / 99.1% cached

    $21.5K20.3%
  3. 03
    Claude Opus 4.8

    Anthropic / 20.9B tokens / 99.1% cached

    $18.2K17.2%
  4. 04
    Claude Opus 4.7

    Anthropic / 13.9B tokens / 99.1% cached

    $11.2K10.5%
  5. 05
    Claude Opus 4.5

    Anthropic / 12.7B tokens / 99.1% cached

    $10.2K9.6%
Show remaining 14 modelsHide remaining models06—19
  1. 06
    Gpt 5.4

    OpenAI / 8.8B tokens / 94.0% cached

    $7.7K7.3%
  2. 07
    Claude Fable 5

    Anthropic / 5B tokens / 99.5% cached

    $7.4K7.0%
  3. 08
    Gpt 5.6 Sol

    OpenAI / 2.9B tokens / 97.2% cached

    $4.7K4.5%
  4. 09
    Gpt 5.3 Codex

    OpenAI / 1.8B tokens / 93.9% cached

    $1.2K1.1%
  5. 10
    Claude Sonnet 4

    Anthropic / 2.4B tokens / 99.1% cached

    $1.2K1.1%
  6. 11
    Gpt 5.2

    OpenAI / 317.5M tokens / 96.2% cached

    $194.060.2%
  7. 12
    Gpt 5.3 Codex Spark

    OpenAI / 252.8M tokens / 93.0% cached

    $173.200.2%
  8. 13
    Gpt 5.2 Codex

    OpenAI / 208.7M tokens / 94.3% cached

    $155.700.1%
  9. 14
    Gpt 5.6 Terra

    OpenAI / 145M tokens / 96.6% cached

    $116.940.1%
  10. 15
    Gpt 5.1 Codex Max

    OpenAI / 70.7M tokens / 71.4% cached

    $71.65<0.1%
  11. 16
    Claude Haiku 4.5

    Anthropic / 199.5M tokens / 99.4% cached

    $37.82<0.1%
  12. 17
    Gpt 5.6 Luna

    OpenAI / 38M tokens / 94.7% cached

    $17.33<0.1%
  13. 18
    Gpt 5.1 Codex Mini

    OpenAI / 4.9M tokens / 83.0% cached

    $0.73<0.1%
  14. 19
    Claude Sonnet 5

    Anthropic / 1.4M tokens / 99.6% cached

    $0.57<0.1%

Not an invoice. This is a standard/global API list-price equivalent; subscription credits, negotiated discounts, and actual card charges are not represented. We use Claude Max and ChatGPT Pro subscriptions, along with usage-based credits.

Section II / Applied trials

Experiment Workstreams

8 workstreams / chronological priority
  1. W1REC-01

    Applied experiment / observed

    AI product and platform architecture

    Jul 2025 — Jul 2026
    Summary
    Designed and hardened an AI-native SaaS platform spanning agent execution, business workflows, shared data, and operational governance.

    Observed outcomes

    1. Connected agent, content, brand, customer, file, and data workflows through shared domain contracts and service boundaries.
    2. Strengthened database models, SDK and API boundaries, approval records, audit paths, and production safeguards.
    3. Used staged releases, targeted tests, and production-readiness reviews to move broad platform changes into operation.

    Index terms

    • AI agents
    • SaaS
    • API design
    • data modeling
    • governance
  2. W2REC-02

    Applied experiment / observed

    Agent developer workflows

    Feb 2026 — Jul 2026
    Summary
    Built reusable infrastructure that makes agent-assisted development easier to install, observe, recover, and extend across tools.

    Observed outcomes

    1. Moved an agent workspace toward a cloud-first operating model with recoverable conversations, clearer failure states, and installable extensions.
    2. Packaged repeatable CLI, SDK, component, and review workflows as portable skills for multiple coding agents.
    3. Published local routing and traffic-inspection tooling for running multi-service development environments from one terminal.

    Index terms

    • agent tooling
    • developer experience
    • CLI
    • workflow
    • observability
  3. W3REC-03

    Applied experiment / observed

    Knowledge and publishing systems

    Jan 2026 — Jul 2026
    Summary
    Developed technical publishing and guided-learning surfaces that keep dense knowledge current, navigable, and useful in practice.

    Observed outcomes

    1. Maintained evolving handbooks for agent tools and operating practices with explicit freshness, search, and indexing policies.
    2. Improved multilingual publishing, content structure, and editorial consistency across long-form technical material.
    3. Turned product research and stories into structured guides, interactive learning paths, and curated reference collections.

    Index terms

    • knowledge systems
    • publishing
    • search
    • localization
    • learning
  4. W4REC-04

    Applied experiment / observed

    Design systems and creative builders

    Jan 2026 — Jul 2026
    Summary
    Created shared editors, builders, UI contracts, and design-governance tools for composing consistent product experiences at speed.

    Observed outcomes

    1. Connected form, theme, screenshot, and asset workflows so output can move cleanly between creative tools.
    2. Improved editor behavior, accessibility, clipboard utilities, rendering performance, and reusable interaction patterns.
    3. Maintained release and documentation practices around shared packages so product surfaces could adopt changes safely.

    Index terms

    • design systems
    • React
    • editors
    • accessibility
    • tooling
  5. W5REC-05

    Applied experiment / observed

    Motion and video production tooling

    Jan 2026 — Jul 2026
    Summary
    Advanced a programmable video workflow from editor input through multi-scene composition, media overlays, and cloud rendering.

    Observed outcomes

    1. Migrated the rendering engine behind a guarded adapter while preserving a clear rollback path.
    2. Added multi-scene sequencing, resolution controls from 720p through 4K, and caption, audio, and music composition.
    3. Hardened runtime isolation and input validation around generated media workloads.

    Index terms

    • video
    • rendering
    • media pipelines
    • cloud
    • runtime safety
  6. W6REC-06

    Applied experiment / observed

    Mobile product delivery

    Jul 2026
    Summary
    Prepared a mobile application shell for production with resilient navigation, messaging, release controls, and failure reporting.

    Observed outcomes

    1. Hardened the web application shell for offline, session, and recovery failure modes.
    2. Implemented deep-link returns, push-token lifecycle handling, remote maintenance, and forced-update controls.
    3. Added production error forwarding and CI gates for type checking, linting, and bundle smoke tests.

    Index terms

    • mobile
    • release engineering
    • deep links
    • push
    • observability
  7. W7REC-07

    Applied experiment / observed

    Data quality and operability

    Jan 2026 — Jul 2026
    Summary
    Improved the reliability of data products by aligning schemas, validation, analytics paths, operational documentation, and recovery procedures.

    Observed outcomes

    1. Reworked analytics and validation paths so generators, scenarios, and data reads share consistent contracts.
    2. Kept schema catalogs and data-flow documentation aligned with production behavior and explicit freshness markers.
    3. Documented health checks, cutover and rollback procedures, and deployment configuration for safer operations.

    Index terms

    • data platforms
    • analytics
    • validation
    • runbooks
    • reliability
  8. W8REC-08

    Applied experiment / observed

    Web discovery and performance

    Jun 2026 — Jul 2026
    Summary
    Built and refined public-facing discovery experiences with stronger information architecture, media quality, rendering performance, and search visibility.

    Observed outcomes

    1. Restructured story and blog routes into clearer product guides with richer metadata and durable URLs.
    2. Expanded curated product information while holding publication to completeness and media-quality checks.
    3. Improved graph rendering, localization coverage, metadata, and SEO behavior across content-heavy pages.

    Index terms

    • content discovery
    • SEO
    • performance
    • information architecture
    • i18n

Archive report is the visual design concept only. This is Eric Han's first-party personal profile; the professional records are factual, not a leaked, investigative, or third-party dossier.