@jeremylongshore
6
week streak
114.1M
token record
18d ago
261
prompt record
18d ago
1244.6M
Total tokens
Build time
4126.9h
Where the tokens go
44% building · 1141.2M classified tokens
Building502.8M44%
Delegating325.0M28%
Verifying164.2M14%
Thinking149.3M13%
25 builds predate classification and aren't included.
2d ago
Swept repository and wired testing gates
cad-dxf-agent
DEEP WORK
First session on cad-dxf-agent
First build tracked for this project
4.9Mtokens
21prompts
47:15:09time
1,698lines
Claude Codeclaude-opus-4-8[1m]Markdown$183.03
mostly building
Deleted 7 merged branches from local state and integrated the audit-harness testing gates into the CI pipeline.
4d ago
Established separate distribution paths for internal vs. public tooling
compile-then-govern
FULL SEND
Top 4% session this week
2088 sessions tracked this week across all builders
3.8Mtokens
10prompts
1:29:49time
349lines
Claude Codeclaude-opus-4-8[1m]Markdown$50.56
mostly thinking
Clarified that internal tooling belongs on a private marketplace, separate from public repository distributions.
4d ago
Audited the repository and cleaned merged branches
clanker
FULL SEND
First session on clanker
First build tracked for this project
1.0Mtokens
13prompts
2:10:20time
38lines
Claude Codeclaude-opus-4-8[1m]Markdown$29.20
mostly building
Fast-forwarded master, validated scaffolding, and reconciled project tracking with GitHub state.
4d ago
Disabled auto-loading of project MCP server
qmd-team-intent-kb
First session on qmd-team-intent-kb
First build tracked for this project
370.7ktokens
6prompts
22:43time
4lines
Claude Codeclaude-opus-4-8[1m]JSON$4.30
mostly building
Created local configuration to prevent a project MCP server from starting automatically.
5d ago
Shipped v0.1.59 after clearing release blockers
agent-governance-plane
DEEP WORK
First session on agent-governance-plane
First build tracked for this project
25.0Mtokens
55prompts
6:42time
3,481lines
Claude Codeclaude-opus-4-8[1m]$509.95
Cleared a dangling tag issue and hardened CI gates to reliably publish v0.1.59.
5d ago
Released v0.11.0 through full quality gates
claude-code-slack-channel
DEEP WORK
First session with opus-4-8[1m]
First build using this model
19.3Mtokens
83prompts
4:25:08time
6,594lines
Claude Codeclaude-opus-4-8[1m]TypeScript$581.87
Shipped v0.11.0 with three PRs through all quality gates; cleaned and stabilized the repository.
5d ago
Fixed production rendering crash and live game updates
braves
FULL SEND
First session with fable-5[1m]
First build using this model
1.2Mtokens
11prompts
43:35time
377lines
Claude Codeclaude-fable-5[1m]TypeScript$15.58
mostly verifying
Fixed a production rendering crash that blocked live games, deployed the fix, and verified all smoke checks pass.
8d ago
Shipped j-rig v2.0.0 with kernel gate migration and validation
Intent Eval Platform
AUTONOMOUS
Biggest session this week
Previous best: 16.2M · +89%
30.6Mtokens
66prompts
48:33:33time
3,256lines
Claude Codeclaude-opus-4-8Markdown$646.24
mostly delegating
Completed kernel gate migration across all packages, released v2.0.0, passed comprehensive review and all CI checks.
8d ago
Solidified schema strategy through expert panel review
Intent Eval Platform
FULL SEND
Top 7% session this week
489 sessions tracked this week across all builders
9.6Mtokens
19prompts
7:51:55time
2,073lines
Claude Codeclaude-opus-4-8Markdown$178.60
balanced mix
Seven experts validated the core schema approach, surfacing plan gaps and critical findings on validator tooling.
10d ago
Locked cost analysis and infrastructure improvements
Intent Eval Platform
MARATHON
4-week shipping streak
28 consecutive days
9.3Mtokens
25prompts
41:03:53time
2,423lines
Claude Codeclaude-opus-4-8YAML$244.62
mostly building
Merged infrastructure updates and finalized plan for cost-driven integration work.
10d ago
Shipped framework for verifiable evidence emission
Intent Eval Platform
MARATHON
Top 11% session this week
514 sessions tracked this week across all builders
6.2Mtokens
27prompts
4:35:52time
4,743lines
Claude Codeclaude-opus-4-8TypeScript$292.23
mostly delegating
Completed intent-eval-core to production standard (99%+ coverage) and locked evidence emission rollout across three evaluation repositories.
10d ago
Planned testing-results portal strategy and evaluation methodology research
Intent Eval Platform
MARATHON
Top 9% session this week
525 sessions tracked this week across all builders
8.1Mtokens
25prompts
17:23:18time
1,381lines
Claude Codeclaude-opus-4-8Markdown$150.44
mostly delegating
Reshaped testing scope, researched evaluation methodologies, and established specs and tracking.