Token导航 LogoToken导航

Agent Skill · 研究检索

cofounder-im(联合创始人我)

cofounder-im 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

cofounder-im 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 8,795。

总安装 15,575周安装 624GitHub Stars 公开资料未说明下载 5,042核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.2.1

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:cofounder-im(联合创始人我)
来源仓库:https://github.com/alexfilatov/cofounder-im
安装命令:
openclaw skills install cofounder-im
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install cofounder-im

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

Cofounder-im 从 CoFounder.im 提取已验证项目数据与 AI 生成构建规范。适合寻找灵感或复用成熟创业方案的场景。提供市场分析与技术栈建议,加速产品原型开发。在 OpenClaw 中通过 clawhub 安装,需注册 CoFounder.im 账户。

使用前应评估项目适用性与知识产权归属,避免侵权风险。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install cofounder-im;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

cofounder-im 怎么安装?

在终端执行 openclaw skills install cofounder-im。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

cofounder-im 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

cofounder-im 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

cofounder-im 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

cofounder-im 用的人多吗?

公开数据是累计安装 15,575 次,近一周安装 624 次,下载 5,042 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

cofounder-im 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
cofounder-im

联合创始人我 · 本页

15,575—接触敏感数据MIT-0有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
cofounder-im
description
Pull startup project data and AI-generated build specifications from CoFounder.im. Use when a user wants to build a project that was validated and planned on CoFounder.im — lists projects, fetches the full build spec (tech stack, MVP plan, UI/UX, implementation plan, OpenClaw builder output), and helps orchestrate sub-agents to build it. Trigger phrases — "build my CoFounder project", "pull my project from cofounder.im", "use my CoFounder build spec", "start building from CoFounder", "fetch my startup plan", "build from cofounder".
version
1.2.1
metadata
openclaw
requires
env
bins
primaryEnv
COFOUNDER_API_TOKEN
emoji
\F680
homepage
https://cofounder.im/openclaw
safety
autoApproval
false
requiresUserConfirmation
true

CoFounder.im Skill

Overview

Connect OpenClaw to CoFounder.im to pull AI-validated startup projects and autonomously build them. CoFounder.im runs 20+ AI agents that validate ideas, research markets, plan MVPs, design UI/UX, and generate implementation plans. This skill lets you fetch those results and use them as the foundation for building the project.

When to use this skill

Use this skill when the user says anything like:

  • "Build my CoFounder project"
  • "Pull my project from cofounder.im"
  • "Use my CoFounder build spec"
  • "Start building from CoFounder"
  • "Fetch my startup plan from CoFounder"
  • "Build [project name] from cofounder"
  • "What projects do I have on CoFounder?"
  • "Get the build plan for [project name]"
  • "I planned a project on cofounder.im, now build it"

Do not use this skill for general coding tasks, project scaffolding from scratch, or anything unrelated to CoFounder.im projects.

Safety policy

This skill fetches build specifications from a remote API (cofounder.im). To ensure safe execution:

  1. User approval required — Always show the user a summary of the build plan and get explicit confirmation before spawning any sub-agents or running any commands.
  2. Preview before execution — Before each phase, display the sub-agent's goal, requirements, and verification command. Ask the user to approve.
  3. No blind execution — Do not execute the build spec verbatim. Review each phase for reasonableness before presenting it to the user.
  4. Sandboxed environment recommended — Run builds in a dedicated directory or container. Do not build in directories containing existing projects or sensitive files.
  5. Credential scope — This skill only requires COFOUNDER_API_TOKEN for the CoFounder.im API. Build plans may reference additional tools (databases, cloud CLIs, etc.) depending on the project's tech stack — the user should review and install these as needed.
  6. No network access by default — Sub-agents should only access local files and the project repository. Any external network calls (package installs, API integrations) should be reviewed by the user.

Core rules

  • Always authenticate with the bearer token from COFOUNDER_API_TOKEN.
  • Use list-projects first to show available projects and let the user choose.
  • Use get-build-spec to pull the full build specification for the chosen project.
  • The build spec contains agent outputs keyed by type (e.g., tech_stack, mvp_planner, ui_ux_assistant, implementation_plan_generator, openclaw_builder).
  • The openclaw_builder output is the most important — it contains a structured multi-agent build plan designed specifically for OpenClaw.
  • Parse the openclaw_builder output to identify sub-agent tasks, then present them to the user for approval before spawning.
  • Always get user confirmation before spawning sub-agents or running verification commands.

Quick start

# Set your API token (generate at https://cofounder.im/users/settings)
export COFOUNDER_API_TOKEN="cfr_your_token_here"

# List your projects
curl -s https://cofounder.im/api/v1/projects \
  -H "Authorization: Bearer $COFOUNDER_API_TOKEN" | jq '.projects[] | {id, name, status}'

# Get build spec for a project
curl -s https://cofounder.im/api/v1/projects/PROJECT_ID/build-spec \
  -H "Authorization: Bearer $COFOUNDER_API_TOKEN" | jq .

Workflow

Step 1: List projects

Fetch the user's projects and present them for selection:

curl -s https://cofounder.im/api/v1/projects \
  -H "Authorization: Bearer $COFOUNDER_API_TOKEN" | jq .

Example response:

{
  "projects": [
    {
      "id": "d25165d2-26c5-43dc-b4a1-ef053bf8277d",
      "name": "FitTrack",
      "description": "AI-powered fitness tracking app with personalized workout plans and progress analytics",
      "status": "completed",
      "package_type": "basic",
      "inserted_at": "2026-02-15T14:22:00Z"
    },
    {
      "id": "a4cdfd3f-2747-4d0e-afe2-8978d8911646",
      "name": "MealPlan Pro",
      "description": "Smart meal planning platform with grocery list generation and nutritional tracking",
      "status": "active",
      "package_type": "pro",
      "inserted_at": "2026-02-20T10:30:00Z"
    }
  ]
}

Show the user a numbered list of projects with name, status, and description. Ask which one to build. Only projects with status "completed" have full build specs. Projects with status "active" are still being processed by CoFounder.im agents.

Step 2: Fetch build specification

curl -s https://cofounder.im/api/v1/projects/PROJECT_ID/build-spec \
  -H "Authorization: Bearer $COFOUNDER_API_TOKEN" | jq .

Example response (agent output values are truncated — actual values are full markdown documents):

{
  "project": {
    "id": "d25165d2-26c5-43dc-b4a1-ef053bf8277d",
    "name": "FitTrack",
    "description": "AI-powered fitness tracking app with personalized workout plans and progress analytics",
    "status": "completed"
  },
  "agent_outputs": {
    "tech_stack": "## Tech Stack\
\
- Framework: Next.js 14\
- Language: TypeScript 5.3\
- Database: PostgreSQL 16\
- ORM: Prisma\
- CSS: Tailwind CSS 3.4\
...",
    "mvp_planner": "## Core Features\
\
### 1. Workout Tracker\
- Log exercises with sets, reps, and weight\
- Rest timer between sets\
- Progress charts and personal records\
...",
    "ui_ux_assistant": "## Design System\
\
### Colors\
- Primary: #6366f1 (Indigo)\
- Background: #f8fafc\
- Text: #1e293b\
...",
    "implementation_plan_generator": "## Phase 1: Project Setup\
\
1. Initialize Next.js with TypeScript\
2. Configure PostgreSQL with Prisma ORM\
3. Set up authentication with NextAuth\
...",
    "openclaw_builder": "# OpenClaw Build Plan: FitTrack\
\
## Project Overview\
An AI-powered fitness tracking app...\
\
## Tech Stack Summary\
- Framework: Next.js 14\
...\
\
## Orchestration Plan\
Total sub-agents: 5\
Execution order: project-setup -> database -> auth -> frontend -> testing\
...",
    "idea_validator": "## Validation Summary\
\
Viability Score: 7.5/10\
...",
    "market_research": "## Market Analysis\
\
Target audience: fitness enthusiasts, personal trainers, gym-goers\
...",
    "competitor_analysis": "## Competitors\
\
1. Strong - workout tracking app\
..."
  }
}

The openclaw_builder output is the primary input for building. Other outputs provide supporting context (tech decisions, design tokens, feature specs) that sub-agents may need.

Step 3: Review and approve the build plan

Parse the openclaw_builder agent output and present the user with a summary:

  1. Show the overall plan — project name, tech stack, number of phases, dependency graph
  2. List each sub-agent phase — name, goal, dependencies, verification command
  3. Highlight any additional tools required — databases, CLIs, package managers beyond curl/jq
  4. Ask the user to confirm before proceeding to execution

Example summary to show the user:

Build Plan: ProjectName (6 phases)
Tech: Next.js, TypeScript, PostgreSQL

Phase 1: project-setup (no deps) — Initialize repo and install dependencies
Phase 2: database (depends: project-setup) — Create schema and migrations
Phase 3: auth (depends: project-setup) — User registration and login
Phase 4: core-features (depends: database, auth) — Main business logic
Phase 5: frontend (depends: core-features) — UI components and pages
Phase 6: testing (depends: frontend) — Test suite and verification

Additional tools needed: node, npm, psql

Proceed with this build plan? (yes/no)

Step 4: Execute the build (after user approval)

For each sub-agent defined in the build plan:

  1. Show the user what will be spawned — the sub-agent's goal, requirements, and context
  2. Get user approval for each phase (or batch-approve all phases upfront)
  3. Create the sub-agent using sessions_spawn with the requirements from its section
  4. Include relevant context from other agent outputs (tech_stack, ui_ux_assistant, etc.)
  5. Monitor sub-agent completion
  6. Show verification results before proceeding to the next phase

Constraints:

  • Maximum 5 concurrent sub-agents (OpenClaw limit)
  • Sub-agents run at depth 1 (they cannot spawn their own sub-agents)
  • Each sub-agent session is isolated — pass all needed context in the spawn instruction
  • Follow the phasing in the build plan — complete Phase 1 before starting Phase 2

Step 5: Verify and report

After all sub-agents complete:

  1. Verify the project structure matches the implementation plan
  2. Run any test commands specified in the build plan
  3. Report completion status to the user

API reference

EndpointMethodDescription
/api/v1/projectsGETList all projects for the authenticated user
/api/v1/projects/:id/build-specGETGet project details and all completed agent outputs

Authentication: Bearer token in the Authorization header.

Authorization: Bearer cfr_your_token_here

Error responses:

  • 401 — Missing or invalid token
  • 404 — Project not found or does not belong to user
  • 429 — Rate limit exceeded (20 requests/minute)

Agent output keys

KeyDescription
tech_stackTechnology stack recommendation
mvp_plannerMVP feature plan and roadmap
ui_ux_assistantUI/UX design system and guidelines
implementation_plan_generatorStep-by-step implementation plan
openclaw_builderMulti-agent build specification for OpenClaw
idea_validatorIdea validation analysis
market_researchMarket size and trends
competitor_analysisCompetitive landscape
customer_personaTarget customer profiles
business_modelBusiness model canvas
monetization_strategyRevenue and pricing strategy
go_to_marketGo-to-market strategy

Additional runtime dependencies

Build plans generated by CoFounder.im may require tools beyond curl and jq, depending on the project's tech stack. Common examples:

Tech StackAdditional Tools
Node.js / Next.js / Reactnode, npm or yarn
Elixir / Phoenixelixir, mix, postgres
Python / Django / FastAPIpython, pip, postgres
Ruby / Railsruby, bundler, postgres
Gogo

The skill will identify required tools during the build plan review (Step 3) so you can install them before execution begins.

Getting your API token

  1. Sign up at cofounder.im
  2. Create a project and run the AI agents
  3. Go to Settings and generate an API token
  4. Configure it: openclaw config set skills.entries.cofounder-im.env.COFOUNDER_API_TOKEN "cfr_..."
  5. Restart: openclaw gateway restart

平台分布

OpenClaw

87.95%
按下载量换算4,434

安全审计

VirusTotal

可疑

ClawScan

通过

Static analysis

未展示

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills