← Back to Skills Marketplace
Replit Agent
by
zhangifonly
· GitHub ↗
· v1.0.1
· MIT-0
153
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install replit-agent
Description
Replit Agent 助手，精通云端 AI 编程、自动部署、数据库集成、全栈应用开发
Usage Guidance
This appears to be an instruction-only Replit assistant (no code or installs). Before installing or using it: (1) understand it currently does not declare any Replit API credentials — if it later asks you to paste tokens or secrets into chat, treat that as sensitive and prefer ephemeral tokens or the official Replit OAuth flow; (2) because the skill has no homepage or published source listed, consider asking the publisher for provenance or source code if you plan to grant account-level access; (3) avoid sharing private keys, passwords, or long-lived tokens directly in messages; (4) if you expect the skill to perform autonomous deployments to your Replit account, confirm how it will obtain and store credentials and review any instructions it provides before executing them. If you want higher assurance, request a version that documents how it integrates with Replit (OAuth endpoints, token storage, or a clear manual workflow).
Capability Analysis
Type: OpenClaw Skill
Name: replit-agent
Version: 1.0.1
The skill bundle consists solely of metadata and informational documentation (SKILL.md) describing the features and capabilities of Replit Agent. There is no executable code, and the instructions provided to the AI agent are purely descriptive and aligned with the stated purpose of being a programming assistant.
Capability Assessment
Purpose & Capability
The name/description (Replit Agent, cloud development, deployment, DB integration) match the SKILL.md content. One minor mismatch: the skill describes automated creation/deployment to Replit but declares no required Replit API token or credential. That can be legitimate for an instruction-only skill (it may only provide guidance and generated code), but it could also mean the skill will prompt users later for tokens if it were to act on their behalf.
Instruction Scope
SKILL.md contains only high-level runtime instructions for the AI (generate projects, deployment guidance, DB integration best practices). It does not tell the agent to read local files, access unrelated environment variables, exfiltrate data, or call unexpected endpoints. No open-ended 'gather whatever context you need' directives are present.
Install Mechanism
No install specification and no code files are present. As an instruction-only skill, it writes nothing to disk and does not fetch external packages — this is the lowest-risk install footprint.
Credentials
The skill requires no environment variables or credentials, which is consistent with an instruction-only helper. However, because it advertises one-click deployment and Secrets management on Replit, a fully autonomous deployment capability would normally need Replit credentials — their absence should be noted and clarified before expecting the skill to act on your account.
Persistence & Privilege
No always:true flag and default autonomous invocation settings. The skill does not request persistent presence, nor does it attempt to modify agent/system-wide configurations. This is standard for a helper skill.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install replit-agent - After installation, invoke the skill by name or use
/replit-agent - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Skill renamed from "Claude Prompt" to "Replit Agent" with a new focus on Replit cloud AI programming.
- Entire SKILL.md content rewritten to describe Replit Agent's capabilities and features.
- Previous focus on Claude prompt engineering replaced with Replit’s agent workflow, cloud deployment, database, and full-stack integration.
- Added comparison table with competing platforms (Bolt.new, Lovable, Cursor).
- Included best practices and practical application scenarios for rapid prototyping, full-stack, API, and learning projects.
v1.0.0
- Initial release of Claude Prompt skill, version 1.0.0
- Provides expert guidance on Claude prompt engineering and system prompt design
- Includes best practices for XML tag structuring, chain-of-thought, role definition, and tool use
- Features advanced techniques such as output control, prefill prompts, long document handling, and tool invocation prompts
- Details common antipatterns and improvement suggestions for better prompt design
- Covers usage scenarios and best practices for maximizing Claude model capabilities
Metadata
Frequently Asked Questions
What is Replit Agent?
Replit Agent 助手，精通云端 AI 编程、自动部署、数据库集成、全栈应用开发. It is an AI Agent Skill for Claude Code / OpenClaw, with 153 downloads so far.
How do I install Replit Agent?
Run "/install replit-agent" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Replit Agent free?
Yes, Replit Agent is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Replit Agent support?
Replit Agent is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Replit Agent?
It is built and maintained by zhangifonly (@zhangifonly); the current version is v1.0.1.
More Skills