You build with AI. Now prove it.
Your best work is invisible.
You spend hours coding with AI, but your best work disappears. Promptbook changes that.
Built iOS target with platform-specific UI and entitlements
1. AI.md Reader
FULL SEND
Top 10% session this week
2454 sessions tracked this week across all builders
1.5Mtokens
19prompts
15:36:55time
891lines
Claude Codeclaude-opus-4-6[1m]Swift$37.06
Extended Pixley Markdown to iOS, enforcing Enhanced-only mode and configuring platform-specific entitlements for both macOS and iOS.
Identified missing search sync in merge append flow
Code
Top 4% build time this week
2456 sessions tracked this week across all builders
317.1ktokens
9prompts
59:54time
Claude Codeclaude-sonnet-4-6$2.96
Found that appending records to existing merges doesn't update the search index like new merges do.
X
@xexyz42· 4hOne-shot debugging on axagents-knowledge
axagents-knowledge
ONE-SHOT
Top 12% build time this week
2457 sessions tracked this week across all builders
142.1ktokens
1prompts
25:45time
35lines
Codexgpt-5.4Python$0.75
25m session running commands and testing. 142K tokens across 1 prompts using gpt-5.4.
X
@xexyz42· 4hDebugged axagents-knowledge
axagents-knowledge
DEEP WORK
33.7x your typical session
4.5M tokens vs typical 133.7K
4.5Mtokens
15prompts
1:17:37time
483lines
Claude Codeclaude-opus-4-6[1m]Python$51.01
1h 17m session running commands and testing. 4.5M tokens across 15 prompts using claude-opus-4-6[1m].
@muzzythinks· 5hConfigured Claude Code performance and caching settings
Mi5
10.5x your typical session
242.1K tokens vs typical 23.1K
242.1ktokens
4prompts
12:46time
5lines
Claude Codeclaude-opus-4-6JSON$1.95
Added environment variables to enable prompt caching and disable adaptive thinking in settings.
Built a complete dockerized video generation and rendering platform
drinkmateandlearn
Top 4% build time this week
2467 sessions tracked this week across all builders
174.2ktokens
3prompts
1:03:29time
635lines
Claude Codeclaude-opus-4-6[1m]Markdown$3.24
Scaffolded backend, frontend, and Remotion services; planned 27 development tasks in a dockerized video system.
Improved mod generation reliability and pattern handling
LAIRD
LOCKED IN
10.5x your typical session
690.6K tokens vs typical 65.9K
690.6ktokens
6prompts
2:59:12time
120lines
Claude Codeclaude-opus-4-5-20251101C#$15.46
Boosted mod generation resilience from 33% to 47% by improving how it handles random and invalid patterns.
Published a three-part guide to building AI agents
portfolio
Top 3% build time this week
2430 sessions tracked this week across all builders
312.6ktokens
9prompts
1:18:24time
92lines
Claude Codeclaude-opus-4-6[1m]$3.53
Shipped three articles on building AI agents from first principles.
Built visual guide to agent architecture patterns
personal
Top 1% build time this week
2394 sessions tracked this week across all builders
111.1ktokens
4prompts
1:58:32time
584lines
Claude Codeclaude-opus-4-6[1m]Markdown$1.48
Created 12 diagrams and Spanish key takeaways documenting agent system design.
Built evaluation system to test AI agent responses
personal
FULL SEND
Top 18% session this week
2357 sessions tracked this week across all builders
811.8ktokens
8prompts
3:53:15time
3,005lines
Claude Codeclaude-sonnet-4-6TypeScript$8.79
Added evaluation infrastructure with 3 datasets and 4 evaluators for agent testing.
Shipped analytics compare page with redesigned stat cards
PROMPTBOOK.GG
FULL SEND
Top 12% session this week
2356 sessions tracked this week across all builders
1.3Mtokens
20prompts
2:37:58time
3,266lines
Claude Codeclaude-opus-4-6[1m]TypeScript$68.54
Launched cross-builder analytics comparison and upgraded the stats dashboard with premium design.
Shipped PWA modernization: offline support and installable app
BLACKJACKRANKED
MARATHON
5.0x your typical session
2.0M tokens vs typical 405.8K
2.0Mtokens
25prompts
2:35:24time
2,494lines
Claude Codeclaude-opus-4-6[1m]JavaScript$88.12
Implemented offline support, installable app, production telemetry, and enhanced mobile interactions.