Agent Plugins Marketplace
← All plugins

thinking-tools

v5.3.0

Structured reasoning, document, issue, and next-session planning skills for Codex.

CodexClaude CodeAgent Plugins9 Skills

By LyaincLicense: MIT0 GitHub starsUpdated 1 hour ago

Directory evidence

Runtimes
Codex, Claude Code, and Agent Plugins
Parsed components
9 skill or MCP entries
Source updated
Sep 26, 2026
Manifest status
Canonical path parsed

The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →

Install thinking-tools for Codex and Claude Code

Installs for the current user
codex plugin marketplace add IchenDEV/agent-plugin-mkt
codex plugin marketplace upgrade agent-plugin-marketplace
codex plugin add thinking-tools@agent-plugin-marketplace

Paste and run these commands in a terminal with Codex. They add and refresh the PluginsMP catalog, then install this plugin.

Compatibility: the page URL and API slug “thinking-tools-4” remain stable.

  • Codex: thinking-tools-4@agent-plugin-marketplace → thinking-tools@agent-plugin-marketplace

The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.

Get the source manually
git clone https://github.com/Lyainc/claude-kit

Clone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is thinking-tools/.

Plugin files

thinking-tools/
├── .codex-plugin/plugin.json
├── .claude-plugin/plugin.json
├── plugin.json
├── skills/adversarial-review/SKILL.md
├── skills/build-spec/SKILL.md
├── skills/diverse-sampling/SKILL.md
├── skills/doc-concretize/SKILL.md
├── skills/doc-polish/SKILL.md
├── skills/expert-panel/SKILL.md
├── skills/issue-raise/SKILL.md
├── skills/next-goal/SKILL.md
└── skills/unknown-discovery/SKILL.md

Included Skills9

adversarial-reviewskills/adversarial-review/SKILL.md

Stress-test claims via adversarial rounds and Survival Score. Runs 1:1 attacker-vs-defender battle. Starts with Steelman. Trigger when user mentions: 반증해줘, 주장 반박, 약점 찾아줘, 논리적 허점 찾아줘, 주장 검증, 살아남을 수 있어?, devil's advocate, adversarial review, claim attack, survival score, steelman and attack. Routing: 합의 도출·다관점은 expert-panel, 맹점 인터뷰는 unknown-discovery.

build-specskills/build-spec/SKILL.md

Crystallize vague ideas into machine-readable Seed specs via Socratic interview and Ambiguity gating. Vendor-neutral, spec-driven workflow. Trigger when user mentions: build-spec, 명세 만들기, 아이디어를 스펙으로, 요구사항 명확화, 아이디어를 명세로, seed 생성, ambiguity gate, requirements crystallize. Routing: 만들 대상이 아직 정해지지 않았거나 위험이 커서 먼저 맹점부터 훑어야 하면 unknown-discovery, "구체화"만 단독이면 doc-concretize, 만들 대상이 정해져 있고 YAML Seed 스펙으로 굳힐 때만 build-spec.

diverse-samplingskills/diverse-sampling/SKILL.md

Generate diverse prose with Verbalized Sampling: explore alternatives or enhance a chosen direction through doc-concretize. Not for factual, single-answer, or code/query work. Trigger when user mentions: 다양한 아이디어, 브레인스토밍, 대안 제시, 창의적 답변, VS 기법으로, diverse ideas, brainstorming, alternatives, verbalized sampling, 글로 발전시켜줘, 더 구체적으로 작성해줘, enhance, 작성 다양성.

doc-concretizeskills/doc-concretize/SKILL.md

Create a new structured Markdown document through recursive concretization and a separate final check; use doc-polish for existing docs and build-spec for YAML specs. Trigger when user mentions: 구체화, 문서화, 체계적 정리, 개념 정리, 아이디어 문서화, 글로 정리, doc-concretize, concretize, "이 개념을 문서로 정리해줘".

doc-polishskills/doc-polish/SKILL.md

Report-only validation of an existing Markdown document: improve structure and cross-check repository-verifiable facts; never rewrite its content. Use doc-concretize for a new document. Trigger when user mentions: 검사해줘, 다듬어줘, 품질 검사, 교정, 다듬기, 문서 사실 확인, 내용이 최신인지, polish, lint, fact check this doc, "이 문서 검사해줘", "README 다듬어줘", "이 설계문서 아직 맞아?".

expert-panelskills/expert-panel/SKILL.md

Facilitate expert panel discussions (thesis-antithesis-synthesis): multiple expert personas debate a decision, each argued in an isolated subagent so positions are independently reasoned (not one model agreeing with itself), optionally grounded in vault citations, converging to consensus + action items. Use when a decision needs several perspectives weighed against each other, not a single answer. Trigger when user mentions: 전문가 토론, 찬반 토론, 다관점 분석, 합의 도출, 트레이드오프 정리, expert panel, multi-perspective review, "전문가 관점에서 검토해줘", "다양한 관점에서 평가해줘". Routing: 1:1 단일 주장 공격은 adversarial-review, 맹점 발견 인터뷰는 unknown-discovery.

issue-raiseskills/issue-raise/SKILL.md

Author and file one GitHub issue from a natural-language request or build-spec Seed, using the repository template, duplicate check, and user approval. Use build-spec first when the requirements need crystallizing. Trigger when user mentions: 이슈 만들어줘, 이슈 저작, 버그 리포트 열어줘, 기능 제안 이슈 올려줘, file an issue, open a github issue, write this up as an issue.

next-goalskills/next-goal/SKILL.md

Choose worthwhile related follow-up work and write a self-contained next-session completion condition. No worthwhile candidate is a valid outcome. Use session-close for the whole closing routine, build-spec for specs, and doc-concretize for documents. Trigger: 완료조건, 다음 세션 목표, START-PROMPT, goal 조건 작성, 다음 작업 정해줘, completion condition, next goal, what should I do next session.

unknown-discoveryskills/unknown-discovery/SKILL.md

Discover Unknown Unknowns through iterative Socratic interviews. Systematically uncover blind spots in projects/plans — the answer comes out as a list of what's unknown, not a spec of what to build (that's build-spec). Has a Quick Discovery Mode (5-7 questions). Trigger when user mentions: 맹점, 놓친 것, 빠진 것, 심층 분석, 인터뷰해줘, 누락된 것, 맹점 검토, blind spot, unknown unknown, "내가 놓치고 있는 게 뭐야?", "이 기획에서 빠진 게 있을까?", 빠르게 맹점만, 간단히 맹점, quick discovery. Routing: 만들 대상이 정해져 있고 명세로 굳혀야 하면 build-spec, 1:1 주장 공격은 adversarial-review, 다관점 합의는 expert-panel.

Plugin manifests3

thinking-tools/.codex-plugin/plugin.json
{
  "name": "thinking-tools",
  "version": "5.3.0",
  "description": "Structured reasoning, document, issue, and next-session planning skills for Codex.",
  "author": {
    "name": "Lyainc",
    "url": "https://github.com/Lyainc"
  },
  "homepage": "https://github.com/Lyainc/claude-kit",
  "repository": "https://github.com/Lyainc/claude-kit",
  "license": "MIT",
  "keywords": [
    "skills",
    "thinking",
    "reasoning",
    "planning",
    "review"
  ],
  "skills": "./skills/",
  "interface": {
    "displayName": "Thinking Tools",
    "shortDescription": "Reasoning and planning skills",
    "longDescription": "Structured reasoning, document, issue authoring, and next-session planning skills for Codex.",
    "developerName": "Lyainc",
    "category": "Productivity",
    "capabilities": [
      "Interactive",
      "Write"
    ],
    "websiteURL": "https://github.com/Lyainc/claude-kit",
    "privacyPolicyURL": "https://github.com/Lyainc/claude-kit/blob/main/LICENSE",
    "termsOfServiceURL": "https://github.com/Lyainc/claude-kit/blob/main/LICENSE",
    "defaultPrompt": [
      "Find the blind spots in this plan",
      "Draft a GitHub issue from this request",
      "Choose the next session goal"
    ],
    "brandColor": "#2563EB"
  }
}
thinking-tools/.claude-plugin/plugin.json
{
  "name": "thinking-tools",
  "version": "5.3.0",
  "description": "Claude Code thinking-tools plugin: 9 skills for structured reasoning, debate, document polish, blind-spot detection, idea crystallization, issue authoring, and next-session completion conditions, plus thinking-facilitator and requirement-gap-reviewer agents.",
  "author": {
    "name": "Lyainc",
    "url": "https://github.com/Lyainc"
  },
  "homepage": "https://github.com/Lyainc/claude-kit",
  "repository": "https://github.com/Lyainc/claude-kit",
  "license": "MIT",
  "keywords": [
    "skills",
    "thinking",
    "expert-panel",
    "doc-concretize",
    "doc-polish",
    "diverse-sampling",
    "unknown-discovery",
    "thinking-facilitator",
    "adversarial-review",
    "build-spec",
    "next-goal",
    "issue-raise",
    "requirement-gap-reviewer"
  ],
  "hooks": {
    "SessionStart": [
      {
        "matcher": "",
        "hooks": [
          {
            "type": "command",
            "command": "\"${CLAUDE_PLUGIN_ROOT}\"/hooks/session-start-welcome.sh",
            "timeout": 5
          }
        ]
      }
    ],
    "PreToolUse": [
      {
        "matcher": "Skill",
        "hooks": [
          {
            "type": "command",
            "command": "\"${CLAUDE_PLUGIN_ROOT}\"/hooks/next-goal-context.sh",
            "timeout": 20
          }
        ]
      },
      {
        "matcher": "Write|Edit",
        "hooks": [
          {
            "type": "command",
            "command": "\"${CLAUDE_PLUGIN_ROOT}\"/hooks/seed-append-guard.sh",
            "timeout": 5
          }
        ]
      }
    ]
  }
}
thinking-tools/plugin.json
{
  "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
  "name": "thinking-tools",
  "version": "5.3.0",
  "description": "Structured reasoning, debate, document polish, blind-spot detection, idea crystallization, issue authoring, and next-session planning skills.",
  "author": {
    "name": "Lyainc",
    "url": "https://github.com/Lyainc"
  },
  "homepage": "https://github.com/Lyainc/claude-kit",
  "repository": "https://github.com/Lyainc/claude-kit",
  "license": "MIT",
  "keywords": [
    "skills",
    "thinking",
    "reasoning",
    "planning",
    "review"
  ]
}

For maintainers

If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.

[thinking-tools on Agent Plugins Marketplace](https://pluginsmp.com/plugins/thinking-tools-4)