AI Agent Skills for Apache APISIX
Agent skills for Apache APISIX are open-source SKILL.md knowledge packs that teach an AI coding agent — such as Claude Code, Cursor, or GitHub Copilot — how to configure and operate a live APISIX gateway. Each skill covers one plugin, recipe, or operator workflow, with the exact commands to run. Install them once, then describe the outcome you want:
"Add key-auth to my
/ordersroute and rate-limit it to 100 requests per minute."
The agent loads the matching skills and runs the real a6 CLI commands against your gateway's Admin API.
The skills below use a6, an open-source (Apache-2.0) command-line tool for Apache APISIX maintained in the api7/a6 repository. a6 wraps the APISIX Admin API; it is not part of an official Apache Software Foundation release. Agents can call the Admin API directly — these skills simply make that ergonomic.
The SKILL.md format is understood by any agent that supports the Agent Skills convention — including Claude Code, Cursor, GitHub Copilot, Windsurf, and OpenCode.
Browse the skills
v1.0.0
a6-persona-developer
Persona skill for API developers building and testing APIs on APISIX using the a6 CLI. Provides decision frameworks for API design, route configuration, plugin selection, testing workflows, local development setup, and CI/CD integration patterns.
v1.0.0
a6-persona-operator
Persona skill for platform operators and DevOps engineers managing APISIX instances using the a6 CLI. Provides decision frameworks for day-to-day operations including deployment, monitoring, troubleshooting, scaling, security hardening, and disaster recovery workflows.
v1.0.0
a6-plugin-ai-content-moderation
Skill for configuring APISIX AI content moderation plugins via the a6 CLI. Covers both ai-aws-content-moderation (AWS Comprehend, request-only) and ai-aliyun-content-moderation (Aliyun, request + response with streaming), toxicity thresholds, category filtering, and integration with ai-proxy.
v1.0.0
a6-plugin-ai-prompt-decorator
Skill for configuring the Apache APISIX ai-prompt-decorator plugin via the a6 CLI. Covers prepending and appending system/user/assistant messages to LLM requests, setting conversation context, enforcing safety guidelines, and combining with ai-proxy and ai-prompt-template in a pipeline.
v1.0.0
a6-plugin-ai-prompt-template
Skill for configuring the Apache APISIX ai-prompt-template plugin via the a6 CLI. Covers defining reusable prompt templates with variable placeholders, enforcing prompt structure, accepting user inputs for specific fields only, and combining with ai-proxy for a complete AI gateway pipeline.
v1.0.0
a6-plugin-ai-proxy
Skill for configuring the Apache APISIX ai-proxy plugin via the a6 CLI. Covers proxying requests to LLM providers (OpenAI, Azure OpenAI, DeepSeek, Anthropic, Gemini, Vertex AI, and more), authentication per provider, model configuration, streaming, logging, and load balancing with ai-proxy-multi.
v1.0.0
a6-plugin-basic-auth
Skill for configuring the Apache APISIX basic-auth plugin via the a6 CLI. Covers HTTP Basic Authentication setup on routes, consumer credential binding with username/password, hide_credentials, anonymous consumer fallback, and common operational patterns.
v1.0.0
a6-plugin-consumer-restriction
Skill for configuring the Apache APISIX consumer-restriction plugin via the a6 CLI. Covers restricting access by consumer name, consumer group ID, service ID, or route ID using whitelist/blacklist modes and per-consumer HTTP method restrictions.
v1.0.0
a6-plugin-cors
Skill for configuring the Apache APISIX cors plugin via the a6 CLI. Covers Cross-Origin Resource Sharing setup on routes, allow_origins, allow_methods, allow_headers, credentials handling, regex origin matching, preflight caching, and common operational patterns.
v1.0.0
a6-plugin-datadog
Skill for configuring the Apache APISIX datadog plugin via the a6 CLI. Covers pushing custom metrics to Datadog via DogStatsD, metric tags, batching, plugin metadata for global DogStatsD server config, and Datadog Agent integration.
v1.0.0
a6-plugin-ext-plugin
Skill for configuring the Apache APISIX external plugin system (ext-plugin-pre-req, ext-plugin-post-req, ext-plugin-post-resp) via the a6 CLI. Covers Plugin Runner architecture, configuration for Go/Java/Python runners, RPC protocol, graceful degradation, and performance considerations.
v1.0.0
a6-plugin-fault-injection
Skill for configuring the Apache APISIX fault-injection plugin via the a6 CLI. Covers injecting delays and HTTP aborts for chaos engineering, percentage-based sampling, conditional injection via vars expressions, custom response headers and body with Nginx variable interpolation.
v1.0.0
a6-plugin-grpc-transcode
Skill for configuring the Apache APISIX grpc-transcode plugin via the a6 CLI. Covers converting RESTful HTTP requests to gRPC, proto file management, pb_option settings for data type conversion, error detail decoding, and common operational patterns.
v1.0.0
a6-plugin-hmac-auth
Skill for configuring the Apache APISIX hmac-auth plugin via the a6 CLI. Covers HMAC signature authentication, consumer credential binding with key_id/secret_key, allowed algorithms, clock skew handling, request body validation, signed headers, and common operational patterns.
v1.0.0
a6-plugin-http-logger
Skill for configuring the Apache APISIX http-logger plugin via the a6 CLI. Covers pushing access logs to HTTP/HTTPS endpoints in batches, custom log formats with NGINX variables, conditional request/response body logging, batch processing tuning, and integration with external logging systems.
v1.0.0
a6-plugin-ip-restriction
Skill for configuring the Apache APISIX ip-restriction plugin via the a6 CLI. Covers IP whitelist/blacklist setup on routes, CIDR range support, IPv4/IPv6, real client IP extraction behind proxies, custom error messages, and common operational patterns.
v1.0.0
a6-plugin-jwt-auth
Skill for configuring the Apache APISIX jwt-auth plugin via the a6 CLI. Covers JWT token authentication, HS256/RS256 algorithm selection, consumer credential binding, token lookup from header/query/cookie, claims handling, clock skew, secret management, and common operational patterns.
v1.0.0
a6-plugin-kafka-logger
Skill for configuring the Apache APISIX kafka-logger plugin via the a6 CLI. Covers pushing access logs to Apache Kafka topics, broker configuration, SASL authentication (PLAIN, SCRAM-SHA-256/512), custom log formats, producer tuning, and batch processing.
v1.0.0
a6-plugin-key-auth
Skill for configuring the Apache APISIX key-auth plugin via the a6 CLI. Covers API key authentication setup on routes, consumer credential binding, key lookup from header/query/cookie, hide_credentials, anonymous consumer fallback, and common operational patterns.
v1.0.0
a6-plugin-limit-count
Skill for configuring the Apache APISIX limit-count plugin via the a6 CLI. Covers fixed-window rate limiting, count/time_window configuration, key types, Redis and Redis-cluster policies for distributed limiting, group-based shared quotas, consumer-level vs route-level limiting, response headers, and common operational patterns.
v1.0.0
a6-plugin-limit-req
Skill for configuring the Apache APISIX limit-req plugin via the a6 CLI. Covers leaky-bucket rate limiting, rate/burst configuration, nodelay behavior, key types, Redis policies for distributed limiting, traffic smoothing, and common operational patterns including combination with limit-count.
v1.0.0
a6-plugin-openid-connect
Skill for configuring the Apache APISIX openid-connect plugin via the a6 CLI. Covers OIDC authorization code flow, bearer token validation, token introspection vs JWKS verification, session management, provider setup for Keycloak/Auth0/Okta, redirect URIs, and common operational patterns.
v1.0.0
a6-plugin-prometheus
Skill for configuring the Apache APISIX prometheus plugin via the a6 CLI. Covers enabling Prometheus metrics export on routes and globally, exposed metrics (HTTP status, latency, bandwidth, upstream health, LLM tokens), custom labels, histogram buckets, and Grafana dashboard integration.
v1.0.0
a6-plugin-proxy-rewrite
Skill for configuring the Apache APISIX proxy-rewrite plugin via the a6 CLI. Covers rewriting request URI, host, method, headers, and scheme before forwarding to upstream. Includes regex URI rewriting, header manipulation, and common operational patterns.
v1.0.0
a6-plugin-redirect
Skill for configuring the Apache APISIX redirect plugin via the a6 CLI. Covers URI redirects, HTTP-to-HTTPS redirection, regex-based URI rewriting, query string handling, and common operational patterns.
v1.0.0
a6-plugin-response-rewrite
Skill for configuring the Apache APISIX response-rewrite plugin via the a6 CLI. Covers rewriting response status codes, headers, and body before returning to clients. Includes conditional execution with vars, regex body filters, base64 body decoding, and common operational patterns.
v1.0.0
a6-plugin-serverless
Skill for configuring the Apache APISIX serverless-pre-function and serverless-post-function plugins via the a6 CLI. Covers inline Lua function execution in configurable request phases, function signature, closure patterns, available Lua APIs, and execution ordering.
v1.0.0
a6-plugin-skywalking
Skill for configuring the Apache APISIX skywalking plugin via the a6 CLI. Covers distributed tracing with Apache SkyWalking OAP, sampling configuration, service topology, and integration with skywalking-logger for trace-log correlation.
v1.0.0
a6-plugin-traffic-split
Skill for configuring the Apache APISIX traffic-split plugin via the a6 CLI. Covers weighted traffic splitting between upstreams with conditional match rules. Includes canary release, blue-green deployment, A/B testing patterns, and common operational workflows.
v1.0.0
a6-plugin-wolf-rbac
Skill for configuring the Apache APISIX wolf-rbac plugin via the a6 CLI. Covers integration with the Wolf RBAC server for role-based access control, token management, login/user-info/change-password API endpoints, permission checking flow, and multi-application setup.
v1.0.0
a6-plugin-zipkin
Skill for configuring the Apache APISIX zipkin plugin via the a6 CLI. Covers distributed tracing with Zipkin, Jaeger, or any Zipkin-compatible collector, B3 propagation headers, sampling, span versions, and trace variable logging.
v1.0.0
a6-recipe-api-versioning
Recipe skill for implementing API versioning strategies using the a6 CLI. Covers URI path versioning with proxy-rewrite, header-based versioning with traffic-split, query parameter versioning, gradual version migration with weighted traffic splitting, and version deprecation with redirect.
v1.0.0
a6-recipe-blue-green
Recipe skill for implementing blue-green deployments using the a6 CLI. Covers creating two upstream environments, switching traffic instantly via route updates or traffic-split plugin, rollback procedures, and config sync workflows for declarative blue-green management.
v1.0.0
a6-recipe-canary
Recipe skill for implementing canary releases using the a6 CLI. Covers gradual traffic shifting with the traffic-split plugin, header-based canary routing, weight adjustment progression, monitoring checkpoints, and full promotion or rollback workflows.
v1.0.0
a6-recipe-circuit-breaker
Recipe skill for implementing circuit breaker patterns using the a6 CLI. Covers the api-breaker plugin for automatic upstream circuit breaking, configuring unhealthy thresholds, healthy recovery, response code classification, and integration with health checks.
v1.0.0
a6-recipe-graphql-proxy
Recipe skill for implementing GraphQL proxying patterns using the a6 CLI. Covers operation-based routing with built-in GraphQL variables, per-operation rate limiting, REST-to-GraphQL conversion with the degraphql plugin, and security patterns for GraphQL APIs.
v1.0.0
a6-recipe-health-check
Recipe skill for configuring upstream health checks using the a6 CLI. Covers active health checks (HTTP probing), passive health checks (response analysis), combining both, configuring healthy/unhealthy thresholds, and monitoring upstream node status.
v1.0.0
a6-recipe-mtls
Recipe skill for configuring mutual TLS (mTLS) using the a6 CLI. Covers SSL certificate management, upstream mTLS to backend services, client certificate verification, and end-to-end mTLS setup from client through APISIX to upstream.
v1.0.0
a6-recipe-multi-tenant
Recipe skill for implementing multi-tenant API gateway patterns using the a6 CLI. Covers tenant isolation via Consumer Groups, host/path/header-based routing, per-tenant rate limiting, context forwarding with proxy-rewrite, and declarative config sync workflows for multi-tenant management.
v1.0.0
a6-shared
Core skill for working with the a6 CLI — the Apache APISIX command-line tool. Provides project conventions, command patterns, architecture overview, and development workflow. Load this skill when working on a6 source code, adding new commands, writing tests, or modifying any a6 component.
Install
Skills are plain Markdown files. Install the pack once and your agent loads the right skill on demand.
Claude Code
# Clone the repo and drop the skills into your agent's skills directory
git clone https://github.com/api7/a6.git
mkdir -p ~/.claude/skills
cp -r a6/skills/* ~/.claude/skills/
Then ask your agent to configure APISIX in natural language — no slash command to memorize.
Other agents
Any agent that reads Markdown skills can use the same files. Clone the repo and point your agent's skills or rules directory at a6/skills/**/SKILL.md (for example, .cursor/rules/ for Cursor).
A one-line installer and a Claude Code plugin package are planned. Track progress in the api7/a6 repository.
Prerequisites
- The
a6CLI installed and on yourPATH(go install github.com/api7/a6/cmd/a6@latest) - A reachable Apache APISIX 3.x instance with Admin API access
- An Admin API key
Example: end to end
A single request like "add key-auth to /orders and rate-limit it to 100 rpm" loads the a6-plugin-key-auth and a6-plugin-limit-count skills. The agent generates the route configuration and applies it with the file-based a6 command:
# route.yaml
uri: /orders
upstream_id: orders-upstream
plugins:
key-auth: {}
limit-count:
count: 100
time_window: 60
key: remote_addr
rejected_code: 429
a6 route update orders -f route.yaml
Frequently asked questions
What are agent skills for Apache APISIX?
Agent skills for Apache APISIX are open-source SKILL.md knowledge packs that teach an AI coding agent how to configure and operate a live APISIX gateway. Each skill covers one plugin, recipe, or operator workflow and the exact commands to run.
Can I use Apache APISIX with Claude Code?
Yes. Install the APISIX agent skills into Claude Code, then describe what you want your gateway to do in natural language — Claude Code loads the matching skill and runs the underlying a6 CLI commands against your APISIX Admin API.
Which AI agents support the APISIX skills?
Any agent that understands the Agent Skills convention, including Claude Code, Cursor, GitHub Copilot, Windsurf, and OpenCode. The skills are plain Markdown files with no runtime or lock-in.
How do I install the APISIX AI agent skills?
Clone the api7/a6 repository and copy its skills into your agent’s skills directory. For Claude Code: git clone https://github.com/api7/a6.git, then cp -r a6/skills/* ~/.claude/skills/.
Is the a6 CLI an official Apache project?
No. a6 is an open-source, Apache-2.0 licensed command-line tool maintained in the api7/a6 repository that wraps the APISIX Admin API. It is not part of an official Apache Software Foundation release.
What can the skills configure?
The catalog spans day-to-day APISIX work: 29 plugin skills (authentication, rate limiting, observability, AI proxying, and more), 8 operational recipes (canary, blue-green, circuit breaking, health checks, mTLS), 2 role personas, and shared CLI conventions.