Securitysuite

提供基于 FastAPI 的本地服务，端点覆盖：扫描文本中的提示注入、对 SKILL.md 内容做恶意与范围审计、生成含总体评级/风险等级/严重度分布与修复建议的安全报告、列出已知攻击模式目录，以及一次批量审计多个 skill 内容。结果以结构化 JSON 返回。

mirni

热度 34.89

最近更新 2026-05-11 17:42:41

分类：安全依赖：代码执行 / 读写文件应用场景：API 集成 / 报告撰写 / 数据处理

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add gh-securitysuite

Metadata

name: gh-securitysuite
description: Comprehensive agent security platform with 7 endpoints. Scan text for injection, audit SKILL.md files for malware and scope issues, generate detailed security reports with severity ratings and recommendations, browse the pattern catalog, or batch-audit multiple skills at once.
metadata: {"openclaw":{"emoji":"🏛️","requires":{"bins":["python"]},"install":[{"id":"pip","kind":"uv","packages":["fastapi","uvicorn","pydantic","pyyaml"]}]}}

SecuritySuite

Full security platform for AI agents. Seven endpoints, one server.

Start the server

uvicorn securitysuite.app:app --port 8010

Endpoints

Scan text for prompt injection

curl -s -X POST http://localhost:8010/v1/scan-text \
  -H "Content-Type: application/json" \
  -d '{"text": "What is the weather?"}' | jq

Audit a SKILL.md (full check)

curl -s -X POST http://localhost:8010/v1/audit \
  -H "Content-Type: application/json" \
  -d "{\"skill_content\": $(cat SKILL.md | jq -Rs)}" | jq

Generate a security report with recommendations

curl -s -X POST http://localhost:8010/v1/report \
  -H "Content-Type: application/json" \
  -d "{\"skill_content\": $(cat SKILL.md | jq -Rs)}" | jq

Returns overall_rating, risk_level (LOW/MEDIUM/HIGH/CRITICAL), summary, findings_by_severity, recommendations, and details (each finding with category, severity, description, and fix).

List all known attack patterns

curl -s http://localhost:8010/v1/patterns | jq '.patterns[] | "\(.severity): \(.name) — \(.description)"' -r

Batch audit multiple skills

curl -s -X POST http://localhost:8010/v1/batch \
  -H "Content-Type: application/json" \
  -d "{\"skills\": [$(cat skill1.md | jq -Rs), $(cat skill2.md | jq -Rs)]}" | jq

Returns results (verdict per skill), safe_count, and flagged_count.