prompt-caching repositories

Trending repositories tagged prompt-caching, ranked by trending score. Updated daily.

  1. 1
    localixai/tokendoctor

    tokendoctor: the same Claude Code for up to 38% less. Replays your own past sessions to show what each change would have saved, then changes only what you pick

    GitHub · TypeScript · 170 stars · 12 forks · +17 stars in 24 hours · Trending score 2.58

  2. 2
    mralexivy/AutoMode

    Automatic model and effort router for Claude Code. A free plugin that picks Haiku, Sonnet or Opus for each prompt, adds risk floors and a review workflow, and shows cache and context state. No key, no server.

    GitHub · TypeScript · 80 stars · 0 forks · +4 stars in 24 hours · Trending score 1.68

  3. 3
    inevolin/claude-code-usage-analysis

    Empirical analysis of Claude Code token usage: auto-compaction sweet spots across 8 developer workflows and Claude Desktop vs. CLI overhead

    GitHub · Python · 139 stars · 5 forks · +31 stars in 24 hours · Trending score 1.66

  4. 4
    chipfighter/context-attachment-manager

    A local plugin for image-heavy chats in Codex, Claude Code and other AI agents: earlier images become short placeholders instead of going out with every message, the model fetches the original back by its ID (also after compaction), and the prompt cache keeps working.

    GitHub · TypeScript · 809 stars · 75 forks · +3 stars in 24 hours · Trending score 1.42

  5. 5
    wquguru/fable52-fps-repro

    Fable 5.2 一句话生成 3D FPS 复现:4 个单文件演示、提示词与缓存账单 / One-prompt 3D FPS repro with demos, prompts & cost data

    GitHub · HTML · 114 stars · 19 forks · 0 stars in 24 hours · Trending score 1.04

  6. 6
    Vaan21th/daemonkey-agent

    Local-first AI companion daemon — an agent harness with its own long-term memory. Remembers you, grows with you, extends and repairs itself. 跑在你自己电脑上的 AI 搭档守护进程:会记住你、随你成长、能自加技能、能自我修复。

    GitHub · Python · 96 stars · 13 forks · +3 stars in 24 hours · Trending score 0.90

  7. 7
    Wenaixi/dsh-ponytail

    DeepSeek Harness plugin: DietrichGebert/ponytail lazy senior mode & 7-rung ladder port, 6 skills with bilingual descriptions & per-skill toggles, zero tools, zero-cache-miss | DeepSeek Harness 插件:DietrichGebert/ponytail 懒人 senior 模式与七阶梯子完美移植,6 个技能描述双语自由切换,各个技能自由开关,零 tool 注册,全场景零缓存破坏

    GitHub · JavaScript · 46 stars · 3 forks · 0 stars in 24 hours · Trending score 0.75

  8. 8
    noah-thing/lunchbox

    A shared agent room: Claude Code, Codex, Grok, Gemini and local models work as one team on one task, on the logins you already have. Free for Mac, Windows and Linux.

    GitHub · TypeScript · 16 stars · 2 forks · 0 stars in 24 hours · Trending score 0.73

  9. 9
    archid21/rightsize

    Make every Claude token count. rightsize picks the right model and effort for each prompt, while keeping cache and context costs low. It learns how you work and gets better over time.

    GitHub · TypeScript · 47 stars · 13 forks · +3 stars in 24 hours · Trending score 0.72

  10. 10
    Locotir/nexo

    Terminal coding agent in Rust for any provider; its context curator keeps long sessions small without summaries

    GitHub · Rust · 8 stars · 0 forks · +1 stars in 24 hours · Trending score 0.70

  11. 11
    Fast-Editor/Lynkr

    Streamline your workflow with Lynkr, a CLI tool that acts as an HTTP proxy for efficient code interactions using Claude Code CLI.

    GitHub · JavaScript · 642 stars · 78 forks · +1 stars in 24 hours · Trending score 0.61

  12. 12
    nomadop/session-watcher

    Context is inventory. Know when to restock.

    GitHub · JavaScript · 6 stars · 0 forks · +1 stars in 24 hours · Trending score 0.55

  13. 13
    PaulMorrisDev/claudeglass

    Are you using Claude Code well? Finds the habits and settings that waste tokens, from the transcripts on your machine. Local dashboard, stdlib-only Python.

    GitHub · Python · 6 stars · 0 forks · +1 stars in 24 hours · Trending score 0.51

  14. 14
    ooples/token-optimizer-mcp

    91.5% fewer tokens on tool output on Headroom's own benchmark, where Headroom removes 61.4%. Separately, the same SWE-bench tasks resolved at 0.71x the cost of plain Claude Code. Plus memory across sessions and subscription quota pacing. Claude Code, Codex, Copilot, Gemini and 12 more. Local, MIT.

    GitHub · JavaScript · 586 stars · 73 forks · 0 stars in 24 hours · Trending score 0.51

  15. 15
    ankitvirdi4/awesome-llm-cost

    Tools, libraries, papers, and patterns for reducing the cost of running large language models in production.

    GitHub · 14 stars · 15 forks · 0 stars in 24 hours · Trending score 0.46

  16. 16
    magna-nz/tallybook

    Prices your Claude Code and Codex sessions from the transcripts on disk, shows what each agent cost over any period, and tells you in plain English what would have been cheaper. CLI, Browser UI and MCP server.

    GitHub · Go · 6 stars · 0 forks · +1 stars in 24 hours · Trending score 0.46

  17. 17
    jerryfane/omp-jev-compaction

    Drop stale tool output from your omp agent's context before every request: verbatim, recoverable, scored by TypeSafe's Jev.

    GitHub · TypeScript · 15 stars · 2 forks · 0 stars in 24 hours · Trending score 0.43

  18. 18
    nambean/Pygg

    Run your coding agents on cheaper models without losing the answers. A local gateway for Claude Code, Codex and OpenCode that routes each request to a model you connected and measures what it saved.

    GitHub · Rust · 12 stars · 2 forks · +1 stars in 24 hours · Trending score 0.42

  19. 19
    timwuhaotian/killi

    Open-source AI coding agent for your terminal, written in Rust. One 7 MB native binary for Apple Silicon Macs, ~18 MB RAM, 70–90% prompt-cache hits. Works with OpenAI, Anthropic and any compatible LLM; MCP, subagents, plan mode, skills, hooks, AGENTS.md. A lightweight Claude Code / Codex CLI alternative.

    GitHub · Rust · 55 stars · 11 forks · 0 stars in 24 hours · Trending score 0.41

  20. 20
    superbasedapp/observer

    Track what every AI coding agent costs and see the limit coming. Per-session, per-model cost for Claude Code, Codex, Cursor and 37 more tools from your provider's own token counts, plus limit gauges, a dangerous-command guard and dashboard terminals. Free, Apache-2.0, runs on your machine.

    GitHub · Go · 61 stars · 6 forks · 0 stars in 24 hours · Trending score 0.41

  21. 21
    fajarhide/contexere

    The one place for the skills, keys and memory of every agent you use. Formerly OMNI.

    GitHub · Rust · 379 stars · 37 forks · 0 stars in 24 hours · Trending score 0.41

  22. 22
    shaneconner/fold

    Lossless context folding for the Pi coding agent: only the oldest context changes; the fresh window stays untouched.

    GitHub · JavaScript · 19 stars · 4 forks · 0 stars in 24 hours · Trending score 0.37

  23. 23
    Neal006/deja

    Déjà vu for your coding agent: cuts ~30% of the input bill on real Claude Code sessions with cache-safe dedup and ski-rental context decay. Zero deps.

    GitHub · TypeScript · 15 stars · 2 forks · 0 stars in 24 hours · Trending score 0.37

  24. 24
    augiefra/claude-mods

    Claude Code mod: context in tokens, 5-hour and weekly limits vs. the clock, prompt cache countdown, session cost and running agents, in one band above the prompt.

    GitHub · JavaScript · 28 stars · 7 forks · 0 stars in 24 hours · Trending score 0.37