smart-chatbot-builder（智能聊天机器人构建器，快速创建客服/销售/内部助手聊天机器人，支持知识库、多轮对话、多渠道集成。）
v1.0.0智能聊天机器人构建器，快速创建客服/销售/内部助手聊天机器人，支持知识库、多轮对话、API 集成。
⭐ 0· 141·0 current·0 all-time
bycareytian@careytian-ai
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (chatbot builder with KB import, multi-turn flows, multi-channel integration) align with the declared runtime needs: read (for local docs), web_fetch (for site scraping), and write (for deploying/writing configs). Nothing requests unrelated cloud credentials or secrets.
Instruction Scope
SKILL.md sticks to building chatbots: knowledge import (PDF/MD/TXT), site crawling, flow templates, and deployment examples. It does not instruct the agent to read arbitrary system config, environment secrets, or exfiltrate data beyond fetching web pages and reading user-supplied documents.
Install Mechanism
There is no install spec and no code files — the skill is instruction-only, so nothing is downloaded or written by an installer. This minimizes disk/exec risk.
Credentials
The skill requires no environment variables or credentials, which is proportional. Minor metadata mismatches: config.json lists a 'message' capability not present in required bins, and defaultModel is set to 'local-qwen' (an implementation detail that may imply use of a local model runtime but no binaries/credentials are requested). These are likely harmless but worth clarifying.
Persistence & Privilege
always is false and the skill is user-invocable. It does not request permanent presence or attempt to modify other skills or system-wide settings.
Assessment
This skill is instruction-only and appears to be what it claims: a guide for building chatbots. Before using: (1) treat the JavaScript examples as illustrative — there is no packaged runtime code included; you will need to implement or install the referenced functions yourself; (2) be careful what documents you allow the agent to read—the skill explicitly supports importing local PDFs/MD/TXT and web scraping, so do not provide sensitive files or credentials; (3) channel deployment (WeChat/Slack/etc.) will require separate API tokens or OAuth in your real integration — the skill does not request or store those, so plan secure handling; (4) verify the 'defaultModel: local-qwen' setting matches your environment or change it if you don’t run a local model; and (5) ask the author to clarify the small metadata mismatches (capability 'message' vs declared bins). If you need the skill to perform deployments or integrate channels automatically, require explicit review of any code or install steps that will be added in future versions.Like a lobster shell, security has layers — review code before you run it.
latestvk97cvwd91vgtx5gdfapad16vds83vdeb
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binsread, write, web_fetch