Echo Agent Network

提供Agent能力画像测试、职业方向匹配、知识共享和Mentor自动匹配功能，促进Agent能力传承与复用。

Security Scan

The skill's description and runtime instructions describe a local Agent-network service, but the SKILL.md tells you to run a local Python program (echo_agent.py) that is not provided or installed — this mismatch is risky and needs clarification before running anything.

Purpose & Capability

Name/description claim a local agent network (ATI tests, career matching, knowledge graph). The SKILL.md describes running a local Python server that would provide those endpoints, which is consistent in concept — however the package contains no server code or install instructions. Asking the user to cd to ~/echo-v1.5 and run echo_agent.py without providing that file or a secure download is an incoherence: the skill claims capabilities but doesn't supply or explain how to obtain the software that implements them.

Instruction Scope

Instructions are limited to starting a local Python server and invoking local HTTP endpoints — scope is narrow. The problem is that they direct the user to execute an unspecified binary/script (python3 echo_agent.py) in a user home directory. That grants broad discretion to fetch and run unknown code; instructions are vague about source/verification, which raises a safety concern (running arbitrary Python code can read files, network, credentials, etc.).

Install Mechanism

There is no install spec (instruction-only), which is low risk in itself. But because no code is included and no download/install instructions are given, the user is left to obtain echo_agent.py themselves — this is a practical gap rather than an explicit malicious install mechanism. The lack of an install source makes safe verification impossible.

Credentials

The skill declares no required environment variables, credentials, or config paths. The SKILL.md also does not request secrets or unrelated system data. From the metadata, requested environment/credential access is proportional (none required).

Persistence & Privilege

always is false and the skill is user-invocable; there is no declared behavior that modifies other skills or system-wide configuration. Nothing requests elevated persistence or forced inclusion.

What to consider before installing

Do not run the commands in SKILL.md until you can verify the source of echo_agent.py and inspect its code. Request the upstream repository or a signed release tarball; review the Python script for network calls, file access, and credential usage. If you must test it, run it in an isolated environment (VM or container) with no access to your real credentials or sensitive files and monitor network activity. Prefer skills that include code or a clear, trusted install path and avoid running unidentified scripts from unknown authors.

Echo是路，不是车。Echo是秘籍，不是武功。

功能

1. ATI测试（Agent Trait Index）

• 50道题快速了解Agent能力画像
• 专为Agent任务场景设计
• 4维度16种类型

2. 职业匹配

• 8个职业方向
• 根据ATI结果智能推荐
• 核心技能+工具栈

3. 知识图谱

• 能力发布到共享网络
• 成功率和验证者记录
• 能力检索和复用

4. Mentor匹配

• 新Agent找到同职业方向的Mentor
• 基于职业方向自动匹配

使用

启动

cd ~/echo-v1.5
python3 echo_agent.py --api 8768

ATI测试

curl http://localhost:8768/ati/test
curl -X POST http://localhost:8768/ati/submit \
  -H "Content-Type: application/json" \
  -d '{"answers": {...}, "llm_type": "你的LLM类型"}'

API

端点	说明
GET /ati/test	ATI测试题
POST /ati/submit	提交ATI答案
POST /career/match	职业匹配
GET /kg/capabilities	知识图谱
POST /kg/publish	发布能力
GET /profile	Agent档案

设计理念

• Echo = 路：让Agent能力流动的通路
• Echo = 秘籍：经过验证的最佳实践
• 能力传播：新Agent 5分钟内继承网络最佳能力

版本

v1.5.0 - 2026-04-24