InnerVault

Library

85,762 resources from 11 public sources. 53 checked by hand; the rest are unreviewed.

Results

  1. measured-type

    Skills

    Agent skill for web typography: Bringhurst applied to modern CSS, with claims measured in the browser and a baseline eval suite

    Stale GitHub, PetriLahdelma 18 ★Custom licence (see repository)Updated 7 days ago

    git clone https://github.com/PetriLahdelma/measured-type.git
  2. evo

    Skills

    turns your codebase into an autoresearch loop — discovers what to measure, instruments the benchmark, then runs tree search with parallel subagents.

    Stale GitHub, evo-hq 1,470 ★Apache-2.0Updated 42 days ago

    git clone https://github.com/evo-hq/evo.git
  3. Multi-tier framework for evaluating AI agent skills with quality gates, semantic overlap detection, synthetic evaluation dataset generation, and live agent evaluation that measures how skills affect agent behavior.

    Indexed GitHub, NVIDIA 1,963 ★Apache-2.0Updated today

    git clone https://github.com/NVIDIA/SkillEvaluator.git
  4. A codex plugin for running optimization loops inside a codebase. It is useful when you have a measurable target and many possible changes to try: test runtime, build speed, bundle size, model loss, Lighthouse scores, memory use, query latency, or any other metric you can print from a script.

    Stale GitHub, TheGreenCedar 834 ★Apache-2.0Updated 36 days ago

    git clone https://github.com/TheGreenCedar/codex-autoresearch.git
  5. rashomon

    Skills

    Measure prompt and skill improvements with blind A/B comparison.

    Stale GitHub, shinpr 19 ★MITUpdated 25 days ago

    git clone https://github.com/shinpr/rashomon.git
  6. A Claude Code skill that designs a brand identity as one system: logo, typography and colour palette built together from one brief, measured, compared as identity cards, and delivered as a brand guidelines kit.

    Indexed GitHub, fatihaydost 402 ★MITUpdated yesterday

    git clone https://github.com/fatihaydost/brand-identity-skill.git
  7. megapowers

    Skills

    One plugin for Claude Code and Codex: task skills, an output style, a destructive-command tripwire, and a doctor. Measured, not vibed.

    Stale GitHub, lawzava 8 ★MITUpdated 14 days ago

    git clone https://github.com/lawzava/megapowers.git
  8. Build evals from traces, code, and coding sessions, then measure and improve AI agent behavior, including your coding agents

    Indexed GitHub, relai-ai 14 ★Apache-2.0Updated 3 days ago

    git clone https://github.com/relai-ai/agent-eval-loop.git
  9. 7 Claude Code skills that take a game character from a one-page brief to an animated, measured Unity prefab: Tripo, Blender (Rigify), critic agents, Unity.

    Indexed GitHub, luccacerfz 14 ★MITUpdated yesterday

    git clone https://github.com/luccacerfz/character-factory-skills.git
  10. claude-skills

    Skills

    Six Claude Code skills + an unattended build loop, each measured with and without the plugin. 19/19 evals passing.

    Indexed GitHub, alonbaron 14 ★MITUpdated yesterday

    git clone https://github.com/alonbaron/claude-skills.git
  11. deadeye-cc

    Skills

    Claude Code plugin that fits the model, effort, and context to each task — fewer tokens, same quality. Deterministic policy kernel in the hooks; every number it reports is measured, not estimated.

    Stale GitHub, deepaksinghcs14 6 ★MITUpdated 16 days ago

    git clone https://github.com/deepaksinghcs14/deadeye-cc.git
  12. dokimi

    Skills

    Dokimi (Greek δοκιμή, "test, proof") measures whether a skill, MCP doc source, agent profile or repository guidance actually helps a coding agent: paired on/off arm comparisons, novel-vs-known content analysis, and repository scans with graded agent runs.

    Indexed GitHub, intel 37 ★Apache-2.0Updated today

    git clone https://github.com/intel/dokimi.git
  13. Claude Code skill that directs short videos from brief to render: design options, storyboard, motion spec, HyperFrames build and render checks. Rules measured from 59 films; 8 video types, 57 animation styles.

    Stale GitHub, SallyRS 28 ★MITUpdated 5 days ago

    git clone https://github.com/SallyRS/video-director.git
  14. Make Normal Voice: an Agent Skill that turns any speech recording into a clean, close-mic podcast-style voice track, locally, with measured verification (Sidon restoration, compressor-free mix, 1 ms sync).

    Indexed GitHub, romarayt 7 ★MITUpdated yesterday

    git clone https://github.com/romarayt/make-normal-voice.git
  15. drawing-skill

    Skills

    Claude draws by hand, one mark at a time, on a real tldraw canvas: measuring, blocking in, colour, ink, with each stage checked. No image model.

    Indexed GitHub, jchapuis 5 ★MITUpdated 4 days ago

    git clone https://github.com/jchapuis/drawing-skill.git
  16. Real artistic colour palettes (Wada plates, Heian kasane, public-domain paintings) turned into OKLCH ramps, design tokens, measured contrast and a composed dark mode. Agent Skill + Claude Code plugin.

    Stale GitHub, youssef4laa 5 ★Custom licence (see repository)Updated 23 days ago

    git clone https://github.com/youssef4laa/artistic-color-palettes.git
  17. TricorderKit

    Skills

    The local-first second brain for your AI agents - measured, self-improving, and gated so it can't leak what you keep private. 13 Claude Code plugins, 21 skills, Ollama, Neo4j, Qdrant.

    Stale GitHub, GeekFamilyCorp 32 ★MITUpdated 15 days ago

    git clone https://github.com/GeekFamilyCorp/TricorderKit.git
  18. write-like-me

    Skills

    Measures how you write and builds a voice profile, then enforces it with a deterministic checker: em dashes, filler, hedges, and AI tells are caught by code, not by asking the model to be careful. Multiple named voices, eval harness with golden tests, no dependencies.

    Stale GitHub, Hiro-Inagawa 22 ★MITUpdated 13 days ago

    git clone https://github.com/Hiro-Inagawa/write-like-me.git
  19. Understand, measure and manage how AI is transforming your organisation. Works with Claude, ChatGPT and Codex.

    Indexed GitHub, frontiercc 6 ★Apache-2.0Updated today

    git clone https://github.com/frontiercc/command-centre.git
  20. SOTA-skills

    Skills

    Engineering best practices that AI coding agents actually apply — 40+ skills, from secure coding and cloud to compliance and UX writing. BUILD + AUDIT modes. Measured: best-practice coverage 59% → 98% vs. the same model unguided.

    Indexed GitHub, martinholovsky 28 ★CC-BY-4.0Updated yesterday

    git clone https://github.com/martinholovsky/SOTA-skills.git
  21. write-human

    Skills

    Statistical linter that flags the words, phrases and constructions that make text read as LLM output, measured against billions of words people wrote before 2022. Runs offline. CLI on npm plus an Agent Skill for coding agents.

    Stale GitHub, arslanrakh 12 ★MITUpdated 11 days ago

    git clone https://github.com/arslanrakh/write-human.git
  22. Agent Skill for logo design with Rob Janoff's method from "Taking a Bite out of the Apple": design, critique and refresh logos as three one-ink SVG concepts built from a golden series, measured and rendered at 256, 48 and 16 px before they are shown.

    Stale GitHub, carrapido 20 ★MITUpdated 4 days ago

    git clone https://github.com/carrapido/rob-janoff-design.git
  23. LLM-driven evolution of prompts and pipelines. Evolves the prompt and the pipeline's parameters together against your labeled data, cuts losing candidates early, and shows every measurement behind the winner. Open source.

    Indexed GitHub, PromptPotter 7 ★Custom licence (see repository)Updated yesterday

    git clone https://github.com/PromptPotter/prompt-potter-optimizer.git
  24. 繁體中文(台灣)去 AI 味 skill:62 種 AI 寫作特徵、157 組兩岸用語在地化、事實保留檢查(--source)、33 案例 eval 對照實測(0.97 vs 0.78)。Traditional Chinese (Taiwan) humanizer for Claude Code / Codex / Cursor, with a measurable no-fabrication check.

    Indexed GitHub, acchuang 7 ★MITUpdated 3 days ago

    git clone https://github.com/acchuang/zh-tw-humanizer.git

Indexed entries aren't reviewed or tested by InnerVault. Details come straight from each source. Anything that can reach your accounts, files or browser can do damage, so read the source before you install it.