TraceGuard Pro
TraceGuard 的商业增值层 —— 面向生产环境的 AI Agent 运行时护栏。
开源核心(MIT)按「工具名 + 静态风险」做判断。TraceGuard Pro 补上受监管团队和生产环境 真正会付费的确定性能力——全部在进程内运行,无网络、无模型调用。
| 能力 | 做什么 | 版本 |
|---|---|---|
| 参数级敏感信息扫描 | 拦截参数里含密钥 / PII 的调用(api_key、password、private_key…) |
Pro |
| 死循环拦截 | 同一个工具 + 参数被反复重试时阻断 | Pro |
| Token 预算 | 单次会话超过 Token 上限时阻断 | Pro |
| 审计导出 | 把 SHA-256 审计链导出为可独立复核的 JSONL | Pro |
| 行业策略包 | 金融 / 医疗 / DevOps 的敏感词 + 预算预设 | Pro |
| 人工审批流 | MEDIUM 风险改为要求人工审批,而不是静默放行 | Team |
| 席位报表 | 按操作者聚合的审计视图 | Team |
| 签名导出 / 自定义包 / SLA | — | Enterprise |
安全设计
- 绝不发起网络请求。 授权码是 Ed25519 非对称签名,本地用公钥验签。
- 源码公开也伪造不出授权码。 客户端只含公钥;只有卖家的私钥能签出有效码。
- 机器绑定(可选)。 一份授权可锁定到某一台机器。
- 完整性自检(可选)。 检测包文件是否被改动。
- 拦截路径无模型调用。 判断是确定性代码。
安装
pip install ai-traceguard traceguard-pro
激活与查看状态
traceguard-pro machine-id # 打印本机机器码(机器绑定用)
traceguard-pro activate TGP1.... # 激活
traceguard-pro status # 查看授权状态
traceguard-pro report audit.db # 导出可校验的审计 JSONL
使用
from traceguard_pro import ProGuard
guard = ProGuard(
policy="policies/base.yaml",
pro_policy="policies-pro/pro.yaml",
license_key="TGP1....", # 离线授权码
mode="strict",
)
result = guard.check("send_email", {"body": "my password=hunter2"})
if not result.allowed:
raise PermissionError(result.reason) # 被 payload_scan 拦下
未持有效授权码时,ProGuard 优雅降级为开源核心行为,方便先试后买。
签发授权码(卖家专用)
私钥是 Ed25519,只存在卖家侧(本地文件 / 私有仓库 Secret),绝不进客户端。
# 本地
python tools/issue_license.py --plan pro --seats 1 --days 365
python tools/issue_license.py --plan team --seats 10 --machine 26e28ebdd786310b
python tools/issue_license.py --plan enterprise --seats 25 --perpetual
# 或在 GitHub 网页 / 手机 App:Actions → Issue TraceGuard Pro License → Run workflow
- 私钥来源:
--key-file或环境变量TRACEGUARD_PRO_PRIVATE_KEY(base64 或 PEM)。 - 打包加固(可选):见
docs/packaging-locked.md。 - 完整性清单:
python tools/freeze_manifest.py
Metadata
Release files for traceguard-pro 0.4.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| traceguard_pro-0.4.0.tar.gz | 25.0 kB | Without provenance | Details for traceguard_pro-0.4.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| traceguard_pro-0.4.0-py3-none-any.whl | Python 3 | none | any | Details for traceguard_pro-0.4.0-py3-none-any.whl |
Total release size: 46.1 kB
Release files / traceguard_pro-0.4.0.tar.gz
| Download URL | traceguard_pro-0.4.0.tar.gz |
|---|---|
| Size | 25.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
b2d737ddb02ba0459185dcb94a3d1adf32706b6d51ffc4e081ea0d7f9beaf57a
|
|
BLAKE2b-256 checksum How to use checksums |
a6bf394b45b98109dea28a3382589bfe3a106aa96783cd3d4d72c52892c17722
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.2
|
Release files / traceguard_pro-0.4.0-py3-none-any.whl
| Download URL | traceguard_pro-0.4.0-py3-none-any.whl |
|---|---|
| Size | 21.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
009a36720c12e4a63564274f51b5960cca0d054bfa7d31dcec68d9bf596ee14f
|
|
BLAKE2b-256 checksum How to use checksums |
45d2c7b17f1c2ed10a966a3e6acde9ff6bdae46180d9254f2b1f00e5f3a044e9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.2
|
Security for traceguard-pro 0.4.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.