Agent skills that run json, that ask for Write

Skills declare what they need inconsistently: 58.8% name the languages they run, 11.2% list the tools they ask for, and 4.6% give themselves a category. Filtering narrows to the ones that said so, never to the ones that are.

1. agent-architecture-auditaffaan-m/ECC

   Full-stack diagnostic for agent and LLM applications. Audits the 12-layer agent stack for wrapper regression, memory pollution, tool discipline failures, hidden repair loops, and r

   pluginbashjsonReadWriteEdit
2. bunqueueegeominotti/bunqueue

   Use bunqueue job queue library - Queue, Worker, Bunqueue (simple mode), FlowProducer, cron, DLQ, embedded and TCP modes

   pluginbashjsontypescriptReadGrepGlob+3 files
3. generate-translationspayloadcms/payload

   Use when new translation keys are added to packages to generate new translations strings

   claude-codejsonWriteBash(date:*)Bash(mkdir -p *)
4. dev-pipelinenexus-substrate/nexus-agents

   Multi-agent development pipeline (Orchestrator + workers + consensus vote). Use when the user asks to "build a feature", "fix a bug with the pipeline", or "run the dev pipeline". T

   pluginjsonReadWriteEdit
5. last30daysmvanhorn/last30days-skill

   Research what people actually say about any topic in the last 30 days. Pulls posts and engagement from Reddit, X, YouTube, TikTok, Hacker News, Polymarket, GitHub, and the web. Inc

   pluginbashjsonBashReadWrite+1 files
6. iwsdk-ui-panelfacebook/immersive-web-sdk

   Create, place, preview, debug, and connect manifest-backed IWSDK UIKitML assets. Use for spatial panels, browser HUDs, UIKitML layout, fonts, editor previews, or runtime element be

   claude-codebashjsontypescriptReadEditWrite
7. bat-adhochomeassistant-ai/ha-mcp

   Run bot acceptance tests to validate MCP tools work correctly from a real AI agent's perspective. Use when testing PRs, detecting regressions, or verifying tool changes end-to-end

   claude-codebashjsonBashReadWrite
8. opencli-browserjackwener/OpenCLI

   Use when an agent needs to drive a real Chrome window via opencli — inspect a page, fill forms, click through logged-in flows, or extract data ad-hoc. Covers the selector-first tar

   otherbashjsonBash(opencli:*)ReadEdit
9. helm-updatemaximhq/bifrost

   Apply helm chart updates for Bifrost. Detects config.schema.json changes since the last helm release, applies user-requested changes, updates values.yaml / values.schema.json / _he

   claude-codebashgo-templatejsonmarkdownReadGrepGlob
10. docs-writermaximhq/bifrost

    Write, update, and review Mintlify MDX documentation for Bifrost features. Explores the full codebase (UI, Go backend, config schema), validates config.json examples, places screen

    claude-codebashjsonmarkdownmdxReadGrepGlob
11. printing-press-publishmvanhorn/cli-printing-press

    Publish a generated CLI to the printing-press-library repo

    pluginbashjsonmarkdownBashReadWrite
12. printing-press-amendmvanhorn/cli-printing-press

    Amend a published CLI from one of two input sources: (1) dogfood mode mines the active Claude Code session transcript for friction (missing flags, hand- rolled API payloads, silent

    pluginbashgojsonyamlBashReadWrite
13. ai-image-creatorcentminmod/my-claude-code-setup

    Generate, edit-from-reference, or analyze images with AI via OpenRouter (gemini, geminipro, riverflow, flux2, seedream, gpt5, gpt5.4; Cloudflare AI Gateway BYOK). Also analyze a vi

    claude-codebashjsonBashReadWrite+2 files
14. figure-specwanshuiyin/Auto-claude-code-research-in-sleep

    Generate deterministic publication-quality architecture, workflow, and pipeline diagrams from structured JSON (FigureSpec) into editable SVG. Use when user says "架构图", "workflow 图"

    otherbashjsonBash(*)ReadWrite+1 files
15. proof-checkerwanshuiyin/Auto-claude-code-research-in-sleep

    Rigorous mathematical proof verification and fixing workflow. Reads a LaTeX proof, identifies gaps via cross-model review (external reviewer backend, ultra reasoning), fixes each g

    otherbashjsonmarkdownBash(*)ReadGrep+3 files
16. research-refinewanshuiyin/Auto-claude-code-research-in-sleep

    Turn a vague research direction into a problem-anchored, elegant, frontier-aware, implementation-oriented method plan via iterative GPT-6-Astra review. Use when the user says "refi

    otherjsonmarkdownBash(*)ReadWrite+3 files
17. auto-paper-improvement-loopwanshuiyin/Auto-claude-code-research-in-sleep

    Autonomously improve a generated paper via GPT-6-Astra xhigh review → implement fixes → recompile, for 2 rounds. Use when user says "改论文", "improve paper", "论文润色循环", "auto improve"

    otherbashjsonmarkdownyamlBash(*)ReadWrite+1 files
18. gemini-searchwanshuiyin/Auto-claude-code-research-in-sleep

    Search research papers via Gemini for broad literature discovery. Use when user says "gemini search", "gemini papers", "search with gemini", or wants AI-powered literature discover

    otherbashjsonBash(*)ReadWrite
19. paper-slideswanshuiyin/Auto-claude-code-research-in-sleep

    Generate conference presentation slides (beamer LaTeX → PDF + editable PPTX) from a compiled paper, with speaker notes and full talk script. Use when user says "做PPT", "做幻灯片", "mak

    otherbashjsonlatexmarkdownBash(*)ReadWrite
20. slides-polishwanshuiyin/Auto-claude-code-research-in-sleep

    Per-page Codex review + targeted python-pptx / Beamer fixes for academic talk slides. Use AFTER /paper-slides (or any externally generated PPTX/Beamer) when the deck looks 'mostly

    otherjsonlatexpythonBash(*)ReadWrite
21. vast-gpuwanshuiyin/Auto-claude-code-research-in-sleep

    Rent, manage, and destroy GPU instances on vast.ai. Use when user says "rent gpu", "vast.ai", "rent a server", "cloud gpu", or needs on-demand GPU without owning hardware.

    otherbashjsonmarkdownBash(*)ReadWrite
22. kill-argumentwanshuiyin/Auto-claude-code-research-in-sleep

    Two-thread adversarial review: a fresh reviewer constructs the strongest 200-word rejection memo, then a second fresh reviewer defends the paper point-by-point and surfaces still-u

    otherbashjsonmarkdownBash(*)ReadWrite+2 files
23. citation-auditwanshuiyin/Auto-claude-code-research-in-sleep

    Zero-context verification that every bibliographic entry in the paper is real, correctly attributed, and used in a context the cited paper actually supports — catching hallucinated

    otherbashjsonmarkdownBash(*)ReadGrep+1 files
24. patent-pipelinewanshuiyin/Auto-claude-code-research-in-sleep

    Full patent drafting pipeline from invention description to jurisdiction-formatted filing documents. Supports CN (CNIPA), US (USPTO), EP (EPO). Supports invention patents and utili

    otherjsonmarkdownBash(*)ReadWrite
25. paper-claim-auditwanshuiyin/Auto-claude-code-research-in-sleep

    Zero-context verification that every number, comparison, and scope claim in the paper matches raw result files. Uses a fresh cross-model reviewer with NO prior context to prevent c

    otherjsonmarkdownBash(*)ReadWrite+1 files
26. idea-creatorwanshuiyin/Auto-claude-code-research-in-sleep

    Generate and rank research ideas given a broad direction. Use when user says "找idea", "brainstorm ideas", "generate research ideas", "what can we work on", or wants to explore a re

    otherbashjsonmarkdownBash(*)ReadWrite+9 files
27. dse-loopwanshuiyin/Auto-claude-code-research-in-sleep

    Autonomous design space exploration loop for computer architecture and EDA. Runs a program, analyzes results, tunes parameters, and iterates until objective is met or timeout. Use

    otherjsonmarkdownBash(*)ReadGrep+1 files
28. experiment-auditwanshuiyin/Auto-claude-code-research-in-sleep

    Audit experiment integrity before claiming results. Uses cross-model review (external reviewer backend) to check for fake ground truth, score normalization fraud, phantom results,

    otherjsonmarkdownBash(*)ReadWrite+1 files
29. qzcliwanshuiyin/Auto-claude-code-research-in-sleep

    Manage GPU compute jobs on the Qizhi (启智) platform using qzcli — a kubectl-style CLI tool. Use when user says "qzcli", "启智平台", "submit job", "stop job", "查计算组", "avail", "list jobs

    otherbashjsonBash(*)ReadWrite
30. paper-writingwanshuiyin/Auto-claude-code-research-in-sleep

    Workflow 3: Full paper writing pipeline that goes from a narrative report to a polished, submission-ready PDF. Use when user says "写论文全流程", "write paper pipeline", "从报告到PDF", "pape

    otherbashjsonmarkdownBash(*)ReadWrite+5 files
31. auto-review-loop-llmwanshuiyin/Auto-claude-code-research-in-sleep

    Autonomous research review loop using any OpenAI-compatible LLM API. Configure via llm-chat MCP server or environment variables. Trigger with "auto review loop llm" or "llm review"

    otherbashjsonmarkdownBash(*)ReadGrep+4 files
32. grant-proposalwanshuiyin/Auto-claude-code-research-in-sleep

    Draft a structured grant proposal from research ideas and literature. Supports KAKENHI (Japan), NSF (US), NSFC (China, including 面上/青年/优青/杰青/海外优青/重点), ERC (EU), DFG (Germany), SNSF

    otherbashjsonmarkdownBash(*)ReadWrite+3 files
33. auto-review-loop-minimaxwanshuiyin/Auto-claude-code-research-in-sleep

    Autonomous multi-round research review loop using MiniMax API. Use when you want to use MiniMax instead of Codex MCP for external review. Trigger with "auto review loop minimax" or

    otherbashjsonmarkdownBash(*)ReadGrep+4 files
34. interview-cheatsheetwanshuiyin/Auto-claude-code-research-in-sleep

    Generate a long-form Chinese interview-prep cheat sheet on a specific ML/LLM topic — formulas with derivations, from-scratch PyTorch code, comparison tables, and 25 高频面试题 (L1 必会 /

    otherbashjsonBash(*)ReadWrite
35. auto-review-loopwanshuiyin/Auto-claude-code-research-in-sleep

    Autonomous multi-round research review loop. In Copilot CLI it defaults to the native complementary rubber-duck subagent with host-event model evidence; elsewhere it uses Codex, wh

    otherbashjsonjsonlmarkdownBash(*)ReadGrep+4 files
36. model-updateBenedictKing/ccx

    新增或更新 CCX 项目中的模型注册技能。适用于 Claude 与非 Claude 模型，覆盖能力表、路由白名单、基准映射、代码生成与验证。

    claude-codebashjsonBashReadWrite
37. upstream-checkBenedictKing/ccx

    检查 Claude Code / Codex 上游版本变更，对比本地版本，识别协议/工具/用法相关更新并追加 TODO 提醒

    claude-codebashjsonmarkdownBashReadWrite
38. generate-gcnpablo-albaladejo/kaiord

    Generate a Garmin Connect (.gcn) workout file from a natural language description. Use when the user describes a training session, workout, or exercise plan and wants a .gcn file.

    claude-codejsonWriteBash+3 files
39. shadcn-uiExplosiveCoderflome/AI-Novel-Writing-Assistant

    Expert guidance for integrating and building applications with shadcn/ui components, including component discovery, installation, customization, and best practices.

    claude-codebashcssjsontypescriptshadcn*:*mcp_shadcn*Read
40. creating-claude-hooksAgentWorkforce/relay

    Use when creating or publishing Claude Code hooks - covers executable format, event types, JSON I/O, exit codes, security requirements, and PRPM package structure

    claude-codebashjsonmarkdowntypescriptReadWriteEdit
41. customise-workflowanombyte93/prd-taskmaster

    Customise the prd-taskmaster plugin workflow via curated brainstorm questions. The AI asks, the user answers in plain English, and the skill writes their preferences to .atlas-ai/c

    pluginbashjsonReadWriteEdit
42. shadcn-uicodedogQBY/ReadAny

    Provides complete shadcn/ui component library patterns including installation, configuration, and implementation of accessible React components. Use when setting up shadcn/ui, inst

    claude-codebashcssjavascriptjsonReadWriteBash
43. skill-federationskill-federation/skill-federation

    Consult the current state of practice before doing expert work — search a federated catalog of vetted agent skills, read the ones that matter as field notes, install only what you'

    pluginbashjsonBashReadWrite+1 files
44. digital-health-clinical-asr-evalNVIDIA/skills

    Stage 3 of Clinical ASR Flywheel. Score a NeMo manifest, produce the five-section KER leaderboard (by-ipa_source diagnostic). Not for ASR auth (/riva-asr).

    pluginjsonReadWriteBash+2 files
45. survey-generatorrohitg00/pro-workflow

    Compile a structured literature survey on any AI/ML topic. Agent curates a research bundle (taxonomy + sections + bibliography of real papers) from a public anchor resource, then a

    pluginbashjsonReadWriteBash
46. agentica-sdkparcadei/Continuous-Claude-v3

    Build Python agents with Agentica SDK - @agentic decorator, spawn(), persistence, MCP integration

    claude-codejsonpythonBashReadWrite
47. skill-developerparcadei/Continuous-Claude-v3

    Meta-skill for creating and managing Claude Code skills

    claude-codebashjsonmarkdownpythonBashReadWrite
48. compound-learningsparcadei/Continuous-Claude-v3

    Transform session learnings into permanent capabilities (skills, rules, agents). Use when asked to "improve setup", "learn from sessions", "compound learnings", or "what patterns s

    claude-codebashjsonmarkdownReadGlobGrep
49. aris-vast-gpuOpenLAIR/dr-claw

    Rent, manage, and destroy GPU instances on vast.ai. Use when user says "rent gpu", "vast.ai", "rent a server", "cloud gpu", or needs on-demand GPU without owning hardware.

    otherbashjsonmarkdownBash(*)ReadWrite
50. aris-paper-slidesOpenLAIR/dr-claw

    Generate conference presentation slides (beamer LaTeX → PDF + editable PPTX) from a compiled paper, with speaker notes and full talk script. Use when user says "做PPT", "做幻灯片", "mak

    otherbashjsonlatexmarkdownBash(*)ReadWrite
51. aris-paper-posterOpenLAIR/dr-claw

    Generate a conference poster (article + tcbposter LaTeX → A0/A1 PDF + editable PPTX + SVG) from a compiled paper. Use when user says "做海报", "制作海报", "conference poster", "make poste

    otherbashjsonlatexmarkdownBash(*)ReadWrite
52. aris-auto-paper-improvement-loopOpenLAIR/dr-claw

    Autonomously improve a generated paper via GPT-5.4 xhigh review → implement fixes → recompile, for 2 rounds. Use when user says "改论文", "improve paper", "论文润色循环", "auto improve", or

    otherbashjsonmarkdownBash(*)ReadWrite
53. aris-research-refineOpenLAIR/dr-claw

    Turn a vague research direction into a problem-anchored, elegant, frontier-aware, implementation-oriented method plan via iterative GPT-5.4 review. Use when the user says "refine m

    otherjsonmarkdownBash(*)ReadWrite
54. aris-auto-review-loopOpenLAIR/dr-claw

    Autonomous multi-round research review loop. Repeatedly reviews via Codex MCP, implements fixes, and re-reviews until positive assessment or max rounds reached. Use when user says

    otherbashjsonmarkdownBash(*)ReadGrep
55. aris-grant-proposalOpenLAIR/dr-claw

    Draft a structured grant proposal from research ideas and literature. Supports KAKENHI (Japan), NSF (US), NSFC (China, including 面上/青年/优青/杰青/海外优青/重点), ERC (EU), DFG (Germany), SNSF

    otherjsonmarkdownBash(*)ReadWrite
56. inno-grant-proposalOpenLAIR/dr-claw

    Help professors and researchers write, revise, adapt, and polish grant proposals for US agencies (NSF, NIH, DOE, DARPA, NASA) and Chinese agencies (NSFC 国自然). Use this skill whenev

    otherbashjsonReadWriteEdit
57. aris-dse-loopOpenLAIR/dr-claw

    Autonomous design space exploration loop for computer architecture and EDA. Runs a program, analyzes results, tunes parameters, and iterates until objective is met or timeout. Use

    otherjsonmarkdownBash(*)ReadGrep
58. stable-releaseanthony-chaudhary/dos-kernel

    Promote an already-shipped rolling release (vX.Y.Z) of the DOS kernel to a named stable channel — gated on a green kernel suite + a green third-party CI run on the candidate + a cl

    claude-codebashjsonmarkdownReadEditGrep
59. dynamicww-w-ai/bkit-claude-code

    Fullstack development with bkend.ai BaaS — authentication, database, API integration. Triggers: fullstack, BaaS, login, signup, database, web app

    pluginbashjsontypescriptReadWriteEdit
60. btwww-w-ai/bkit-claude-code

    By-The-Way: capture improvement ideas the moment they surface, without derailing the task in progress. Triggers: /bkit:btw, /bkit:btw list, /bkit:btw analyze, /bkit:btw promote, /b

    pluginjsonReadWriteEdit

Browse all 69,865 skills →200 at a time, grouped by owner. The list above is the skills with the highest standing; this is every one of them.

The whole corpus is one call away through the search API, or ask a question of it on the index itself.