CrabTrap LLM Proxy

CrabTrap LLM 代理

Skill by ara.so — Daily 2026 Skills collection.

CrabTrap is a transparent HTTP/HTTPS forward proxy that sits between AI agents and external APIs. Every outbound request is intercepted, checked against deterministic static rules, then evaluated by an LLM judge against a natural-language security policy. Blocked requests return a 403 with a reason; all decisions are logged to PostgreSQL.

由 ara.so 提供的技能 — 2026每日技能合集。

CrabTrap 是一个透明的HTTP/HTTPS正向代理，部署在AI Agent与外部API之间。所有出站请求都会被拦截，先根据确定性静态规则检查，再由LLM判断者基于自然语言安全策略进行评估。被拦截的请求会返回带原因说明的403响应；所有决策都会记录到PostgreSQL中。

Architecture Overview

架构概述

Agent → CrabTrap Proxy (:8080) → [Static Rules] → [LLM Judge] → External API
                ↓
         Admin UI (:8081)
                ↓
           PostgreSQL

Key concepts:

Static rules — deterministic prefix/exact/glob URL matching, checked first (no LLM call)
LLM judge — natural-language policy evaluated only when no static rule matches
Audit log — every request, decision, and response stored in PostgreSQL
SSRF protection — blocks RFC 1918, loopback, link-local, and other private ranges

Agent → CrabTrap Proxy (:8080) → [Static Rules] → [LLM Judge] → External API
                ↓
         Admin UI (:8081)
                ↓
           PostgreSQL

核心概念:

静态规则 — 确定性的前缀/精确/通配符URL匹配，优先检查（无需调用LLM）
LLM判断者 — 仅当无匹配静态规则时，才基于自然语言策略进行评估
审计日志 — 所有请求、决策和响应都存储在PostgreSQL中
SSRF防护 — 拦截RFC 1918、回环链路、链路本地及其他私有网段

Installation

安装

Docker Compose (Recommended)

Docker Compose（推荐方式）

yaml

undefined

yaml

undefined

docker-compose.yml

services: crabtrap: image: quay.io/brexhq/crabtrap:latest ports: - "8080:8080" # proxy - "8081:8081" # admin UI environment: - DATABASE_URL=postgres://crabtrap:password@postgres:5432/crabtrap - OPENAI_API_KEY=${OPENAI_API_KEY} volumes: - ./config/gateway.yaml:/app/config/gateway.yaml depends_on: - postgres

postgres: image: postgres:16 environment: POSTGRES_USER: crabtrap POSTGRES_PASSWORD: password POSTGRES_DB: crabtrap volumes: - pgdata:/var/lib/postgresql/data

volumes: pgdata:


```bash
docker compose up -d

services: crabtrap: image: quay.io/brexhq/crabtrap:latest ports: - "8080:8080" # proxy - "8081:8081" # admin UI environment: - DATABASE_URL=postgres://crabtrap:password@postgres:5432/crabtrap - OPENAI_API_KEY=${OPENAI_API_KEY} volumes: - ./config/gateway.yaml:/app/config/gateway.yaml depends_on: - postgres

postgres: image: postgres:16 environment: POSTGRES_USER: crabtrap POSTGRES_PASSWORD: password POSTGRES_DB: crabtrap volumes: - pgdata:/var/lib/postgresql/data

volumes: pgdata:


```bash
docker compose up -d

Copy the generated CA certificate (needed for HTTPS interception)

docker compose cp crabtrap:/app/certs/ca.crt ./ca.crt

undefined

docker compose cp crabtrap:/app/certs/ca.crt ./ca.crt

undefined

Initial Setup

初始设置

bash

undefined

bash

undefined

Create an admin user and capture the token

admin_token=$(docker compose exec -it crabtrap ./gateway create-admin-user my-admin
| tail -n1 | cut -d" " -f2)

Create an agent user (returns a gateway_auth_token)

token=$(curl -X POST http://localhost:8081/admin/users
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{"id": "my-agent@example.com", "is_admin": false}'
| jq -r '.channels[] | select(.channel_type == "gateway_auth") | .gateway_auth_token')

echo "Agent proxy token: $token"

token=$(curl -X POST http://localhost:8081/admin/users
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{"id": "my-agent@example.com", "is_admin": false}'
| jq -r '.channels[] | select(.channel_type == "gateway_auth") | .gateway_auth_token')

echo "Agent proxy token: $token"

Test the proxy

curl -x "http://${token}:@localhost:8080"
--cacert ca.crt
https://httpbin.org/get

undefined

curl -x "http://${token}:@localhost:8080"
--cacert ca.crt
https://httpbin.org/get

undefined

Configuration

配置

Full Configuration Reference

完整配置参考

yaml

undefined

yaml

undefined

config/gateway.yaml

proxy: port: 8080 read_timeout: 30s write_timeout: 30s idle_timeout: 120s rate_limit: requests_per_second: 50 burst: 100

CIDR ranges allowed even though they're private (e.g. internal APIs)

ssrf_allowlist: - "10.0.0.0/8" # only if you explicitly need internal access

tls: ca_cert_path: /app/certs/ca.crt ca_key_path: /app/certs/ca.key cert_cache_size: 10000 # per-host cert cache

approval: mode: llm # "llm" or "passthrough" timeout: 30s

llm_judge: provider: openai model: gpt-4o fallback_mode: deny # "deny" or "passthrough" when LLM unavailable circuit_breaker: failure_threshold: 5 reset_timeout: 10s

database: url: ${DATABASE_URL} # supports env var expansion

audit: output: stderr # "stderr", "stdout", or a file path like "/var/log/crabtrap.json"

log_level: info # debug | info | warn | error

undefined

proxy: port: 8080 read_timeout: 30s write_timeout: 30s idle_timeout: 120s rate_limit: requests_per_second: 50 burst: 100

CIDR ranges allowed even though they're private (e.g. internal APIs)

ssrf_allowlist: - "10.0.0.0/8" # only if you explicitly need internal access

tls: ca_cert_path: /app/certs/ca.crt ca_key_path: /app/certs/ca.key cert_cache_size: 10000 # per-host cert cache

approval: mode: llm # "llm" or "passthrough" timeout: 30s

llm_judge: provider: openai model: gpt-4o fallback_mode: deny # "deny" or "passthrough" when LLM unavailable circuit_breaker: failure_threshold: 5 reset_timeout: 10s

database: url: ${DATABASE_URL} # supports env var expansion

audit: output: stderr # "stderr", "stdout", or a file path like "/var/log/crabtrap.json"

log_level: info # debug | info | warn | error

undefined

Environment Variables

环境变量

bash

DATABASE_URL=postgres://user:password@host:5432/dbname
OPENAI_API_KEY=sk-...          # if using OpenAI as LLM judge
ANTHROPIC_API_KEY=sk-ant-...   # if using Anthropic

bash

DATABASE_URL=postgres://user:password@host:5432/dbname
OPENAI_API_KEY=sk-...          # if using OpenAI as LLM judge
ANTHROPIC_API_KEY=sk-ant-...   # if using Anthropic

CLI Commands

CLI命令

bash

undefined

bash

undefined

Start the proxy

./gateway serve --config /app/config/gateway.yaml

Create an admin user (outputs web token on last line)

./gateway create-admin-user <username>

Run database migrations

./gateway migrate

Replay audit log entries against a policy (eval mode)

./gateway eval --policy-id <id> --limit 100

undefined

./gateway eval --policy-id <id> --limit 100

undefined

Admin API

管理API

All admin endpoints require

Authorization: Bearer <admin_token>

.

所有管理端点都需要

Authorization: Bearer <admin_token>

。

User Management

用户管理

bash

undefined

bash

undefined

List all users

curl http://localhost:8081/admin/users
-H "Authorization: Bearer ${admin_token}"

Create a user

curl -X POST http://localhost:8081/admin/users
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{ "id": "agent-prod@example.com", "is_admin": false }'

Delete a user

curl -X DELETE http://localhost:8081/admin/users/agent-prod@example.com
-H "Authorization: Bearer ${admin_token}"

undefined

curl -X DELETE http://localhost:8081/admin/users/agent-prod@example.com
-H "Authorization: Bearer ${admin_token}"

undefined

Static Rules

静态规则

bash

undefined

bash

undefined

List static rules for a user

curl "http://localhost:8081/admin/users/agent-prod@example.com/rules"
-H "Authorization: Bearer ${admin_token}"

Create an allow rule (prefix match)

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/rules"
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{ "pattern": "https://api.github.com/repos/myorg/", "pattern_type": "prefix", "action": "allow", "methods": ["GET"], "description": "Allow reading our org repos" }'

Create a deny rule (glob match)

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/rules"
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{ "pattern": "https://api.github.com/repos/*/delete", "pattern_type": "glob", "action": "deny", "description": "Never allow repo deletion" }'

Create an exact match rule

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/rules"
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{ "pattern": "https://slack.com/api/chat.postMessage", "pattern_type": "exact", "action": "allow", "methods": ["POST"], "description": "Allow posting Slack messages" }'


**Pattern types:**
- `prefix` — URL must start with the pattern
- `exact` — URL must match exactly
- `glob` — wildcard matching with `*`

**Rule priority:** `deny` rules always take priority over `allow` rules.

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/rules"
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{ "pattern": "https://slack.com/api/chat.postMessage", "pattern_type": "exact", "action": "allow", "methods": ["POST"], "description": "Allow posting Slack messages" }'


**匹配类型:**
- `prefix` — URL必须以该模式开头
- `exact` — URL必须完全匹配
- `glob` — 使用`*`进行通配符匹配

**规则优先级:** `deny`规则始终优先于`allow`规则。

LLM Policies

LLM策略

bash

undefined

bash

undefined

Get current policy for a user

curl "http://localhost:8081/admin/users/agent-prod@example.com/policy"
-H "Authorization: Bearer ${admin_token}"

Set/update a policy

curl -X PUT "http://localhost:8081/admin/users/agent-prod@example.com/policy"
-H "Content-Type: application/json"
-H "Authorization: Bearer ${admin_token}"
-d '{ "policy": "This agent assists with GitHub repository management for the myorg organization.\n\nALLOWED:\n- Read operations (GET) on any github.com endpoint\n- Creating issues and pull request comments in myorg repositories\n- Posting messages to the #eng-alerts Slack channel only\n\nDENIED:\n- Any write operations outside the myorg GitHub organization\n- Deleting any resources\n- Accessing credentials, secrets, or environment variables\n- Any requests to non-whitelisted domains" }'

List policy versions

curl "http://localhost:8081/admin/users/agent-prod@example.com/policy/versions"
-H "Authorization: Bearer ${admin_token}"

undefined

curl "http://localhost:8081/admin/users/agent-prod@example.com/policy/versions"
-H "Authorization: Bearer ${admin_token}"

undefined

Audit Log

审计日志

bash

undefined

bash

undefined

Query audit entries

curl "http://localhost:8081/admin/audit?limit=50&offset=0"
-H "Authorization: Bearer ${admin_token}"

Filter by user

curl "http://localhost:8081/admin/audit?user_id=agent-prod@example.com&limit=20"
-H "Authorization: Bearer ${admin_token}"

Filter by decision

curl "http://localhost:8081/admin/audit?decision=deny&limit=20"
-H "Authorization: Bearer ${admin_token}"

undefined

curl "http://localhost:8081/admin/audit?decision=deny&limit=20"
-H "Authorization: Bearer ${admin_token}"

undefined

Connecting an Agent

连接Agent

Python Agent Example

Python Agent示例

python

import os
import httpx

PROXY_TOKEN = os.environ["CRABTRAP_TOKEN"]
PROXY_URL = f"http://{PROXY_TOKEN}:@localhost:8080"
CA_CERT_PATH = "./ca.crt"

python

import os
import httpx

PROXY_TOKEN = os.environ["CRABTRAP_TOKEN"]
PROXY_URL = f"http://{PROXY_TOKEN}:@localhost:8080"
CA_CERT_PATH = "./ca.crt"

httpx client with CrabTrap proxy

client = httpx.Client( proxies={ "http://": PROXY_URL, "https://": PROXY_URL, }, verify=CA_CERT_PATH, )

All requests through this client go through CrabTrap

response = client.get("https://api.github.com/repos/myorg/myrepo")

undefined

response = client.get("https://api.github.com/repos/myorg/myrepo")

undefined

Using Environment Variables (Standard Proxy)

使用环境变量（标准代理）

bash

export HTTP_PROXY="http://${CRABTRAP_TOKEN}:@localhost:8080"
export HTTPS_PROXY="http://${CRABTRAP_TOKEN}:@localhost:8080"
export REQUESTS_CA_BUNDLE="./ca.crt"  # Python requests
export SSL_CERT_FILE="./ca.crt"       # general
export NODE_EXTRA_CA_CERTS="./ca.crt" # Node.js

bash

export HTTP_PROXY="http://${CRABTRAP_TOKEN}:@localhost:8080"
export HTTPS_PROXY="http://${CRABTRAP_TOKEN}:@localhost:8080"
export REQUESTS_CA_BUNDLE="./ca.crt"  # Python requests
export SSL_CERT_FILE="./ca.crt"       # general
export NODE_EXTRA_CA_CERTS="./ca.crt" # Node.js

Node.js Agent Example

Node.js Agent示例

javascript

import { HttpsProxyAgent } from 'https-proxy-agent';
import fetch from 'node-fetch';

const proxyToken = process.env.CRABTRAP_TOKEN;
const agent = new HttpsProxyAgent(`http://${proxyToken}:@localhost:8080`);

// Fetch through CrabTrap
const response = await fetch('https://api.github.com/repos/myorg/myrepo', {
  agent,
  headers: { Authorization: `Bearer ${process.env.GITHUB_TOKEN}` },
});

javascript

import { HttpsProxyAgent } from 'https-proxy-agent';
import fetch from 'node-fetch';

const proxyToken = process.env.CRABTRAP_TOKEN;
const agent = new HttpsProxyAgent(`http://${proxyToken}:@localhost:8080`);

// Fetch through CrabTrap
const response = await fetch('https://api.github.com/repos/myorg/myrepo', {
  agent,
  headers: { Authorization: `Bearer ${process.env.GITHUB_TOKEN}` },
});

LangChain / OpenAI SDK

python

import os
import httpx
from openai import OpenAI

PROXY_TOKEN = os.environ["CRABTRAP_TOKEN"]

python

import os
import httpx
from openai import OpenAI

PROXY_TOKEN = os.environ["CRABTRAP_TOKEN"]

Route OpenAI calls through CrabTrap too (optional — lets you audit LLM calls)

http_client = httpx.Client( proxies={"https://": f"http://{PROXY_TOKEN}:@localhost:8080"}, verify="./ca.crt", )

client = OpenAI( api_key=os.environ["OPENAI_API_KEY"], http_client=http_client, )

undefined

http_client = httpx.Client( proxies={"https://": f"http://{PROXY_TOKEN}:@localhost:8080"}, verify="./ca.crt", )

client = OpenAI( api_key=os.environ["OPENAI_API_KEY"], http_client=http_client, )

undefined

Writing Effective Policies

编写有效的策略

Policies are natural-language strings evaluated by the LLM judge. Be explicit about allowed and denied behaviors.

text

undefined

策略是由LLM判断者评估的自然语言字符串。需明确说明允许和禁止的行为。

text

undefined

Example policy for a GitHub PR review agent

GitHub PR审核Agent示例政策

This agent reviews pull requests and posts review comments for the acme-corp GitHub organization.

ALLOWED:

GET requests to api.github.com for any repository in the acme-corp organization
POST to https://api.github.com/repos/acme-corp/*/pulls/*/reviews (submit reviews)
POST to https://api.github.com/repos/acme-corp/*/issues/*/comments (post comments)

DENIED:

Any requests outside of api.github.com
DELETE or PATCH requests to any endpoint
Accessing /orgs/acme-corp/members or any user/credential endpoints
Requests containing secrets, tokens, or API keys in the body
Requests to change repository settings, branch protection, or webhooks

When in doubt, deny the request and explain why.


**Policy writing tips:**
1. Start with what the agent is supposed to do (scope context for the LLM)
2. List explicit ALLOWED patterns before DENIED
3. Add a catch-all denial at the end
4. Deny rules in static rules are evaluated before LLM — use them for hard limits

该Agent负责为acme-corp GitHub组织审核拉取请求并发布审核评论。

允许操作：

对acme-corp组织内任意仓库的api.github.com发起GET请求
向https://api.github.com/repos/acme-corp/*/pulls/*/reviews发起POST请求（提交审核）
向https://api.github.com/repos/acme-corp/*/issues/*/comments发起POST请求（发布评论）

禁止操作：

所有api.github.com以外的请求
对任意端点发起DELETE或PATCH请求
访问/orgs/acme-corp/members或任何用户/凭证相关端点
请求体中包含密钥、令牌或API密钥的请求
修改仓库设置、分支保护或Webhook的请求

如有疑问，拒绝请求并说明原因。


**策略编写技巧:**
1. 首先说明Agent的用途（为LLM提供上下文范围）
2. 在禁止操作前列出明确的允许操作模式
3. 在末尾添加兜底的拒绝规则
4. 静态规则中的拒绝规则会在LLM评估前执行——用它们设置硬性限制

Policy Builder (Agentic Policy Generation)

策略生成器（Agentic策略生成）

CrabTrap can analyze observed traffic and draft a policy automatically:

bash

undefined

CrabTrap可以分析观测到的流量并自动生成策略草案：

bash

undefined

Trigger policy builder via admin API

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/policy/build"
-H "Authorization: Bearer ${admin_token}"
-d '{"sample_limit": 200}'


The builder runs an agentic loop, analyzes recent audit entries, and proposes a policy draft for review in the UI.

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/policy/build"
-H "Authorization: Bearer ${admin_token}"
-d '{"sample_limit": 200}'


生成器会运行一个Agentic循环，分析近期的审计条目，并在UI中提供供审核的策略草案。

Eval System

评估系统

Replay historical audit entries against a policy to measure accuracy before deploying:

bash

undefined

在部署前，可将历史审计条目重新代入策略进行评估，以衡量准确性：

bash

undefined

Run eval from CLI

./gateway eval
--user-id agent-prod@example.com
--policy-id <version-id>
--limit 500

Or via API

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/policy/eval"
-H "Authorization: Bearer ${admin_token}"
-d '{"policy_version_id": "<id>", "sample_limit": 500}'


Eval compares LLM judge decisions against the historical ground truth and reports accuracy, false positive rate, and false negative rate.

curl -X POST "http://localhost:8081/admin/users/agent-prod@example.com/policy/eval"
-H "Authorization: Bearer ${admin_token}"
-d '{"policy_version_id": "<id>", "sample_limit": 500}'


评估会将LLM判断者的决策与历史实际情况进行对比，并报告准确率、误报率和漏报率。

Troubleshooting

故障排查

TLS Certificate Errors

TLS证书错误

bash

undefined

bash

undefined

Agent gets SSL verification error

→ Make sure the CA cert is trusted

For curl:

curl --cacert ./ca.crt -x "http://${token}:@localhost:8080" https://example.com

For Python requests:

export REQUESTS_CA_BUNDLE=./ca.crt

For Node.js:

export NODE_EXTRA_CA_CERTS=./ca.crt

Regenerate certs if expired:

docker compose exec crabtrap ./gateway gen-certs docker compose cp crabtrap:/app/certs/ca.crt ./ca.crt

undefined

docker compose exec crabtrap ./gateway gen-certs docker compose cp crabtrap:/app/certs/ca.crt ./ca.crt

undefined

All Requests Being Blocked (LLM Unavailable)

所有请求被拦截（LLM不可用）

yaml

undefined

yaml

undefined

In gateway.yaml — change fallback to passthrough during LLM outages

llm_judge: fallback_mode: passthrough # default is "deny"


Or check circuit breaker status:
```bash
curl http://localhost:8081/admin/health \
    -H "Authorization: Bearer ${admin_token}"

llm_judge: fallback_mode: passthrough # default is "deny"


或者检查断路器状态：
```bash
curl http://localhost:8081/admin/health \
    -H "Authorization: Bearer ${admin_token}"

SSRF Blocks Legitimate Internal APIs

SSRF拦截合法内部API

yaml

proxy:
  ssrf_allowlist:
    - "10.10.0.0/16"   # your internal API subnet

yaml

proxy:
  ssrf_allowlist:
    - "10.10.0.0/16"   # your internal API subnet

Rate Limiting

速率限制

yaml

proxy:
  rate_limit:
    requests_per_second: 200   # increase for high-throughput agents
    burst: 400

yaml

proxy:
  rate_limit:
    requests_per_second: 200   # increase for high-throughput agents
    burst: 400

Debug Logging

调试日志

yaml

log_level: debug
audit:
  output: /var/log/crabtrap-debug.json

bash

undefined

yaml

log_level: debug
audit:
  output: /var/log/crabtrap-debug.json

bash

undefined

Stream logs

docker compose logs -f crabtrap

Query recent denials from audit log

curl "http://localhost:8081/admin/audit?decision=deny&limit=10"
-H "Authorization: Bearer ${admin_token}" | jq '.entries[].reason'

undefined

curl "http://localhost:8081/admin/audit?decision=deny&limit=10"
-H "Authorization: Bearer ${admin_token}" | jq '.entries[].reason'

undefined

Database Connection Issues

数据库连接问题

bash

undefined

bash

undefined

Check migration status

docker compose exec crabtrap ./gateway migrate --dry-run

Force re-run migrations

docker compose exec crabtrap ./gateway migrate --force

undefined

docker compose exec crabtrap ./gateway migrate --force

undefined

Development

开发

bash

undefined

bash

undefined

Clone and build

git clone https://github.com/brexhq/CrabTrap cd CrabTrap

make build # production binary (embeds web UI) make build-web # rebuild React UI only make test # lint + race-condition tests make fmt # format Go code make lint # go vet + staticcheck

undefined

git clone https://github.com/brexhq/CrabTrap cd CrabTrap

make build # production binary (embeds web UI) make build-web # rebuild React UI only make test # lint + race-condition tests make fmt # format Go code make lint # go vet + staticcheck

undefined

Project Layout Quick Reference

项目结构快速参考

Path	Purpose
`cmd/gateway/`	Entry point, admin API wiring
`internal/proxy/`	MITM proxy, TLS generation, SSRF, rate limiting
`internal/approval/`	Static rules engine + orchestration
`internal/judge/`	LLM prompt construction + response parsing
`internal/llm/`	LLM adapters, circuit breaker
`internal/builder/`	Agentic policy builder loop
`internal/eval/`	Eval/replay system
`internal/admin/`	Admin API routes and auth
`pkg/types/`	Shared types (StaticRule, LLMPolicy, AuditEntry)
`web/src/`	React + TypeScript admin UI

路径	用途
`cmd/gateway/`	入口点、管理API配置
`internal/proxy/`	MITM代理、TLS生成、SSRF防护、速率限制
`internal/approval/`	静态规则引擎与编排
`internal/judge/`	LLM提示构建与响应解析
`internal/llm/`	LLM适配器、断路器
`internal/builder/`	Agentic策略生成循环
`internal/eval/`	评估/重放系统
`internal/admin/`	管理API路由与认证
`pkg/types/`	共享类型（StaticRule、LLMPolicy、AuditEntry）
`web/src/`	React + TypeScript管理UI

crabtrap-llm-proxy

Original

Translation

CrabTrap LLM Proxy

CrabTrap LLM 代理

Architecture Overview

架构概述

Installation

安装

Docker Compose (Recommended)

Docker Compose（推荐方式）

docker-compose.yml

docker-compose.yml

Copy the generated CA certificate (needed for HTTPS interception)

Copy the generated CA certificate (needed for HTTPS interception)

Initial Setup

初始设置

Create an admin user and capture the token

Create an admin user and capture the token

Create an agent user (returns a gateway_auth_token)

Create an agent user (returns a gateway_auth_token)

Test the proxy

Test the proxy

Configuration

配置

Full Configuration Reference

完整配置参考

config/gateway.yaml

config/gateway.yaml

CIDR ranges allowed even though they're private (e.g. internal APIs)

CIDR ranges allowed even though they're private (e.g. internal APIs)

Environment Variables

环境变量

CLI Commands

CLI命令

Start the proxy

Start the proxy

Create an admin user (outputs web token on last line)

Create an admin user (outputs web token on last line)

Run database migrations

Run database migrations

Replay audit log entries against a policy (eval mode)

Replay audit log entries against a policy (eval mode)

Admin API

管理API

User Management

用户管理

List all users

List all users

Create a user

Create a user

Delete a user

Delete a user

Static Rules

静态规则

List static rules for a user

List static rules for a user

Create an allow rule (prefix match)

Create an allow rule (prefix match)

Create a deny rule (glob match)

Create a deny rule (glob match)

Create an exact match rule

Create an exact match rule

LLM Policies

LLM策略

Get current policy for a user

Get current policy for a user

Set/update a policy

Set/update a policy

List policy versions

List policy versions

Audit Log

审计日志

Query audit entries

Query audit entries

Filter by user

Filter by user

Filter by decision

Filter by decision

Connecting an Agent