Personal project allocation

Attention and agent-resource signals · 14d window since 2026-10-21, unless labeled otherwise.

59 projects · 3,130 paired responses · 2,199 MTok · 38% sub-agent output

Workflow intensity × tokens

Each point is a project. Both axes use log scales; hover a point for exact values.

1101001,00010,000100,000110100 Inquisition groupchat · intensity 55,017 · tokens 88.85 MTok · compute skew 0.72×Harvest decompilation · intensity 15,543 · tokens 26.27 MTok · compute skew 0.40×Panetone · intensity 13,753 · tokens 33.40 MTok · compute skew 0.55×endchat · intensity 13,686 · tokens 5.70 MTok · compute skew 0.09×wsh · intensity 10,535 · tokens 40.85 MTok · compute skew 0.77×FreeTranscribe.org · intensity 8,762 · tokens 116 MTok · compute skew 2.38×Warcraft III agent · intensity 5,797 · tokens 204 MTok · compute skew 5.16×Trading competition · intensity 5,471 · tokens 2.85 MTok · compute skew 0.07×Crimsonland decomp · intensity 5,010 · tokens 27.33 MTok · compute skew 0.75×Concord matching recompiler · intensity 5,001 · tokens 450 MTok · compute skew 12.29×gtube YouTube Music player · intensity 4,278 · tokens 5.11 MTok · compute skew 0.15×Git contributions · intensity 3,604 · tokens 4.72 MTok · compute skew 0.15×Wakterm.org · intensity 3,291 · tokens 104 MTok · compute skew 3.49×gitrep.fyi · intensity 3,079 · tokens 65.04 MTok · compute skew 2.27×Agent Orchestration Process · intensity 2,953 · tokens 55.08 MTok · compute skew 1.96×Automated reverse engineering · intensity 2,915 · tokens 354 MTok · compute skew 12.69×Model prices · intensity 2,755 · tokens 68.23 MTok · compute skew 2.51×Gesso · intensity 2,744 · tokens 30.57 MTok · compute skew 1.13×rev.ng contributions · intensity 2,204 · tokens 45.66 MTok · compute skew 1.88×Personal release management · intensity 2,101 · tokens 6.84 MTok · compute skew 0.29×MihaiCosma.com · intensity 1,679 · tokens 16.64 MTok · compute skew 0.79×bend-bench · intensity 1,495 · tokens 4.57 MTok · compute skew 0.23×clanker-usage · intensity 1,357 · tokens 12.92 MTok · compute skew 0.68×agent-spec-sheet · intensity 1,161 · tokens 4.11 MTok · compute skew 0.23×Polymarket research backtests · intensity 1,139 · tokens 42.25 MTok · compute skew 2.43×Search clanker history · intensity 845 · tokens 10.84 MTok · compute skew 0.72×Clanker analytics · intensity 775 · tokens 20.50 MTok · compute skew 1.43×Infobase · intensity 765 · tokens 72.20 MTok · compute skew 5.07×UFOpedia.fyi · intensity 723 · tokens 52.10 MTok · compute skew 3.77×Personal Project Tracking · intensity 673 · tokens 14.19 MTok · compute skew 1.06×llama.cpp fork · intensity 243 · tokens 4.31 MTok · compute skew 0.54×Bend · intensity 178 · tokens 14.19 MTok · compute skew 2.08×Personal scripts · intensity 173 · tokens 25.46 MTok · compute skew 3.78×Vellum · intensity 99.4 · tokens 9.82 MTok · compute skew 1.93×groot · intensity 83.0 · tokens 12.69 MTok · compute skew 2.73×Zola · intensity 74.9 · tokens 6.23 MTok · compute skew 1.41×Options market-making research · intensity 33.5 · tokens 0.66 MTok · compute skew 0.22×NikoRun app · intensity 3.4 · tokens 0.19 MTok · compute skew 0.21×Aipocalypse.dev · intensity 1.3 · tokens 0.40 MTok · compute skew 0.70× ↖ more compute-led more attention-led ↘ Workflow intensity Tokens (MTok) log-log fit · ρ 0.47

Projects

Metrics use the 14d window unless labeled otherwise.

Project Attention Agent resources Automation
ResponsesYour messages that followed a finished AI turn within one hour, in interactive Claude Code and Codex sessions. They include messages typed in a terminal or sent from a chat on any device, and exclude other chat members and messages from other agents. LatencyMedian time from a finished AI turn to your next message, excluding delays of one hour or more. IntensityActive session time divided by median sub-hour response latency. A session runs from your first message to the last finished AI turn and ends after an hour without a message from you. It rises with longer project sessions and faster supervision cadence. 14d shareShare of total workflow-intensity scores across all measured projects in the 14d reporting window, including private projects omitted from this table. This is a relative engagement index, not a share of clock time.1d shareShare of total workflow-intensity scores across all measured projects in the 1d comparison window, including private projects omitted from this table. Compare it with the main-window share to see where workflow focus has shifted. TokensMillions of total tokens after discounting cache-read tokens by 90%, matching Clanker's normalized resource measure. Compute skewActual normalized tokens divided by the tokens predicted from workflow intensity by the chart's log-log fit. Above 1× is compute-heavy; below 1× is workflow-heavy. Sub-agentsShare of AI output generated by sub-agents or headless executions. AOPShare of AI output generated with aop.
Inquisition groupchat 1,323 16s 55,017 25.5%39.2% 88.85 0.72× 0% 0%
Harvest decompilation 2 1.4m 15,543 7.2%6.4% 26.27 0.40× 0% 0%
Panetone 55 1.1m 13,753 6.4%3.6% 33.40 0.55× 0% 0%
endchat 15 1.2m 13,686 6.4%0.0% 5.70 0.09× 0% 0%
wsh 56 1.5m 10,535 4.9%0.0% 40.85 0.77× 0% 0%
FreeTranscribe.org 107 1.8m 8,762 4.1%2.6% 116 2.38× 0% 0%
Warcraft III agent 270 2.3m 5,797 2.7%6.3% 204 5.16× 2% 2%
Trading competition 5 57s 5,471 2.5%0.0% 2.85 0.07× 0% 0%
Crimsonland decomp 44 2.3m 5,010 2.3%0.0% 27.33 0.75× 0% 0%
Concord matching recompiler 161 2.9m 5,001 2.3%2.5% 450 12.29× 81% 81%
gtube YouTube Music player 19 1.3m 4,278 2.0%0.0% 5.11 0.15× 0% 0%
Git contributions 38 1.6m 3,604 1.7%0.0% 4.72 0.15× 0% 0%
Wakterm.org 106 3.6m 3,291 1.5%0.6% 104 3.49× 0% 0%
gitrep.fyi 86 2.9m 3,079 1.4%2.2% 65.04 2.27× 0% 0%
Agent Orchestration Process 35 5.2m 2,953 1.4%2.9% 55.08 1.96× 0% 0%
Automated reverse engineering 138 4.4m 2,915 1.4%1.5% 354 12.69× 1% 1%
Model prices 58 3.7m 2,755 1.3%0.0% 68.23 2.51× 1% 1%
Gesso 19 4.6m 2,744 1.3%3.5% 30.57 1.13× 0% 0%
rev.ng contributions 109 3.6m 2,204 1.0%0.5% 45.66 1.88× 0% 0%
Personal release management 2 8.9m 2,101 1.0%0.0% 6.84 0.29× 0% 0%
MihaiCosma.com 20 7.6m 1,679 0.8%15.8% 16.64 0.79× 0% 0%
bend-bench 4 4.5m 1,495 0.7%0.0% 4.57 0.23× 0% 0%
clanker-usage 16 9.7m 1,357 0.6%0.0% 12.92 0.68× 0% 0%
agent-spec-sheet 8 5.6m 1,161 0.5%0.0% 4.11 0.23× 0% 0%
Polymarket research backtests 106 1.9m 1,139 0.5%0.1% 42.25 2.43× 6% 6%
Search clanker history 17 8.8m 845 0.4%0.0% 10.84 0.72× 0% 0%
Clanker analytics 9 17.3m 775 0.4%0.0% 20.50 1.43× 0% 0%
Infobase 140 4.3m 765 0.4%1.8% 72.20 5.07× 0% 0%
UFOpedia.fyi 31 2.8m 723 0.3%0.2% 52.10 3.77× 0% 0%
Personal Project Tracking 41 6.1m 673 0.3%0.1% 14.19 1.06× 86% 86%
llama.cpp fork 2 42.3m 243 0.1%0.0% 4.31 0.54× 0% 0%
Bend 18 4.1m 178 0.1%0.0% 14.19 2.08× 0% 0%
Personal scripts 3 51.8m 173 0.1%0.0% 25.46 3.78× 5% 5%
Vellum 9 6.0m 99.4 0.0%0.0% 9.82 1.93× 0% 0%
groot 33 6.7m 83.0 0.0%0.0% 12.69 2.73× 0% 0%
Zola 15 3.2m 74.9 0.0%0.0% 6.23 1.41× 0% 0%
Options market-making research 8 42s 33.5 0.0%0.0% 0.66 0.22× 0% 0%
NikoRun app 1 2.6m 3.4 0.0%0.0% 0.19 0.21× 0% 0%
Aipocalypse.dev 1 2.9m 1.3 0.0%0.0% 0.40 0.70× 0% 0%
Document extraction utilities 0 n/a n/a 0.0%0.0% 94.68 n/a 48% 48%
Harvest oracle 0 n/a n/a 0.0%0.0% 23.79 n/a 98% 98%
hearsay 0 n/a n/a 0.0%0.0% 5.95 n/a 0% 0%
Cheap food optimizer 0 n/a n/a 0.0%0.0% 4.30 n/a 0% 0%
Codex 0 n/a n/a 0.0%0.0% 3.93 n/a 0% 0%
AntennaPod 0 n/a n/a 0.0%0.0% 3.10 n/a 0% 0%
redvsblue.fyi 0 n/a n/a 0.0%0.0% 0.89 n/a 0% 0%
freetranscribe 0 n/a n/a 0.0%0.0% 0.85 n/a 0% 0%
gstream 0 n/a n/a 0.0%0.0% 0.81 n/a 0% 0%
Hardware review 0 n/a n/a 0.0%0.0% 0.58 n/a 0% 0%
BendTorch 0 n/a n/a 0.0%0.0% 0.57 n/a 0% 0%
Debate groupchat 0 n/a n/a 0.0%0.0% 0.48 n/a 0% 0%
DeepSeek Harness 0 n/a n/a 0.0%0.0% 0.34 n/a 0% 0%
code-peon 0 n/a n/a 0.0%0.0% 0.19 n/a 0% 0%
vmaf 0 n/a n/a 0.0%0.0% 0.17 n/a 0% 0%
encode-score 0 n/a n/a 0.0%0.0% 0.08 n/a 0% 0%
Clanker Mono 0 n/a n/a 0.0%0.0% 0.04 n/a 0% 0%
Bend ML 0 n/a n/a 0.0%0.0% 0.03 n/a 0% 0%
ASR GPU inference optimization 0 n/a n/a 0.0%0.0% 0.01 n/a 0% 0%
loophole 0 n/a n/a 0.0%0.0% 0.01 n/a 0% 0%