Skip to content
devhub.
ExploreCompareSaved
Sign in
DevHubBuilt for curious developers. Data from GitHub.
Back to repositories
Public repository / warrenth

ctxcraft

Evaluate and optimize your AI agent context. Save tokens, save money.

View on GitHub
13Stars
0Forks
1Open issues
0Subscribers

Language distribution

Share of source bytes reported by GitHub

Shell100%

Recent activity

Latest commits on the default branch

feat: align with official Claude Code docs (2026) — path-scoped rules, spec frontmatter, unified scoringwarrenth · Jul 27, 2026docs: clarify plugin install must be run per project20250387 · May 4, 2026revert: remove --loop mode (not applicable to static config optimization)warrenth · Mar 27, 2026feat: add AutoResearch-inspired loop mode to READMEwarrenth · Mar 27, 2026feat: add --loop mode to /optimize (inspired by Karpathy's AutoResearch)warrenth · Mar 27, 2026feat: add grade scale (A/A-/B+/B/C/D/F) to evaluate and optimize reportswarrenth · Mar 27, 2026[feat] optimize Strategy 4→6 연쇄 로직 추가 — dedup check, auto-trigger, execution flow 통합warrenth · Mar 27, 2026feat: add locale detection for multi-language report output20250387 · Mar 18, 2026docs: rewrite README in English for global audience20250387 · Mar 17, 2026[feat] 2축 평가 체계 도입 — 품질(universal) + 비용(per plan tier)20250387 · Mar 17, 2026

Project details

Created Mar 2026 Updated Jul 2026 MIT License
ai-agentscluade-codecontext-engineeringdeveloper-toolsharness-engineeringprompt-engineeringtoken-optimization

Contributors

2
warrenth's avatarwarrenth48 commits20250387's avatar202503877 commits