AGENTS.md

1<!--
2 Licensed to the Apache Software Foundation (ASF) under one or more
3 contributor license agreements.  See the NOTICE file distributed with
4 this work for additional information regarding copyright ownership.
5 The ASF licenses this file to You under the Apache License, Version 2.0
6 (the "License"); you may not use this file except in compliance with
7 the License.  You may obtain a copy of the License at
8 
9    http://www.apache.org/licenses/LICENSE-2.0
10 
11 Unless required by applicable law or agreed to in writing, software
12 distributed under the License is distributed on an "AS IS" BASIS,
13 WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 See the License for the specific language governing permissions and
15 limitations under the License.
16-->
17 
18# Agent Guide for Apache Kafka
19 
20This file is read by automated agents (security scanners, code analyzers,
21AI assistants) operating on this repository.
22 
23## Project overview
24 
25Apache Kafka is a distributed event streaming platform. This repo is a **Gradle**
26build in **Java** and **Scala 2.13**.
27 
28- **Supported JDK**: 11+ for `clients`, `generator`, and `streams`; 17+ for other modules.
29- **Build/test JDK**: Java 17 and 25.
30- **Build tool**: Gradle, wrapper `./gradlew` is recommended.
31 
32## Repository layout (high level)
33 
34| Path                                                                                          | Purpose                                                                                            |
35|-----------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------|
36| `clients/`                                                                                    | Producer, consumer, admin client APIs and protocol messages                                        |
37| `core/`                                                                                       | Broker runtime, log, replication, request handling (gradually being removed along with Scala code) |
38| `server/`, `server-common/`                                                                   | Broker/server components and shared server code                                                    |
39| `metadata/`, `raft/`                                                                          | KRaft metadata layer and Raft implementation                                                       |
40| `storage/`, `storage/api/`                                                                    | Log segments, checkpoints, tiered storage APIs                                                     |
41| `group-coordinator/`, `share-coordinator/`, `transaction-coordinator/`, `coordinator-common/` | Coordinators                                                                                       |
42| `connect/`                                                                                    | Kafka Connect (api, runtime, plugins, transforms)                                                  |
43| `streams/`                                                                                    | Kafka Streams (+ Scala, examples, upgrade system tests)                                            |
44| `generator/`                                                                                  | RPC/message code generation                                                                        |
45| `tools/`, `shell/`                                                                            | CLI tools and kafka metadata shell                                                                 |
46| `examples/`                                                                                   | Kafka producer and consumer examples                                                               |
47| `jmh-benchmarks/`                                                                             | Kafka benchmarks tests                                                                             |
48| `trogdor/`                                                                                    | test framework                                                                                     |
49| `tests/`                                                                                      | System test harness (see `tests/README.md`)                                                        |
50| `config/`                                                                                     | Sample broker/controller configuration                                                             |
51| `checkstyle/`                                                                                 | Checkstyle rules and import-control XML                                                            |
52| `docs/`                                                                                       | Documentation sources                                                                              |
53| `docker/`                                                                                     | Docker images, build/test scripts, and examples for JVM and native Kafka           |
54 
55 
56## Build commands
57 
58```bash
59# Compile jars
60./gradlew jar
61 
62# Release tarball (output under core/build/distributions/)
63./gradlew clean releaseTarGz
64 
65# Publish to local Maven (skip signing for local dev)
66./gradlew -PskipSigning=true publishToMavenLocal
67 
68# Regenerate RPC/message classes after branch switches
69./gradlew processMessages processTestMessages
70 
71# List tasks
72./gradlew tasks
73```
74 
75Module-scoped build:
76 
77```bash
78./gradlew core:jar
79./gradlew clients:jar
80./gradlew :streams:testAll
81```
82 
83## Test commands
84 
85```bash
86# All unit + integration tests
87./gradlew test
88 
89# Split suites
90./gradlew unitTest
91./gradlew integrationTest
92 
93# Include flaky tests (marked @Flaky)
94./gradlew test -Pkafka.test.run.flaky=true
95 
96# Re-run without considering up-to-date
97./gradlew test --rerun-tasks
98 
99# Single test class
100./gradlew clients:test --tests RequestResponseTest
101 
102# Single test method
103./gradlew clients:clients-integration-tests:test --tests org.apache.kafka.clients.producer.ProducerFailureHandlingTest.testCannotSendToInternalTopic
104 
105# Coverage (whole project or one module)
106./gradlew reportCoverage -PenableTestCoverage=true -Dorg.gradle.parallel=false
107./gradlew clients:reportCoverage -PenableTestCoverage=true -Dorg.gradle.parallel=false
108```
109 
110**Test retries** (optional): `-PmaxTestRetries=1 -PmaxTestRetryFailures=3`
111 
112**Logging**: adjust `log4j2.yaml` under `<module>/src/test/resources/` for verbose test logs.
113 
114**System tests**: see [tests/README.md](tests/README.md).
115 
116## Code quality
117 
118Run before submitting Java changes:
119 
120```bash
121./gradlew spotlessApply          # fix import order / formatting
122./gradlew checkstyleMain checkstyleTest spotlessCheck
123./gradlew spotbugsMain spotbugsTest -x test   # optional static analysis
124```
125 
126Checkstyle reports: `<module>/build/reports/checkstyle/`.
127 
128## Coding conventions
129 
130- **Commits / PR titles**: Start with `KAFKA-XXXXX`, `MINOR`, or `HOTFIX`. Use `KAFKA-XXXXX` only when there is a valid Jira ticket for the change.
131- **AI-generated contributions**: Follow the [AI-Generated Contributions](CONTRIBUTING.md#ai-generated-contributions) section in `CONTRIBUTING.md` — add a `Co-Authored-By` or `Generated-by` commit trailer for AI-assisted changes.
132- **Public API / KIP**: Changes to public interfaces, wire protocol, configurations, or metrics generally require a [KIP](https://cwiki.apache.org/confluence/display/KAFKA/Kafka+Improvement+Proposals). See the `javadoc` `include` list in [build.gradle](build.gradle) for packages treated as public API.
133- **License header**: New source files need the standard ASF license header (see existing files in the same module).
134- **Checkstyle**: Enforced on main and test sources; rules in `checkstyle/`. Import order is checked via Spotless.
135- **Generated code**: Do not hand-edit generated RPC/message sources. Regenerate with `processMessages` / `processTestMessages`. See [clients/src/main/resources/common/message/README.md](clients/src/main/resources/common/message/README.md).
136- **Scope**: Prefer targeted unit tests in the module you change; integration/system tests for cross-component behavior.
137- **Compatibility**: clients, generator, and streams modules support Java 11+; all other modules require Java 17+. Avoid APIs that break those release levels.
138 
139## Security
140 
141See [SECURITY.md](./SECURITY.md) for how to report a vulnerability and for links
142to the Apache Kafka security model under [docs/security/](./docs/security/).
143 
144Agents that scan this repository should consult `SECURITY.md` and the linked
145security model before reporting issues. In particular, the core model defines
146what is in and out of scope, how reports are classified, and a list of known
147non-findings; each component page adds its own known non-findings.
148 
149## Reference docs
150 
151| Topic                          | Location |
152|--------------------------------|----------|
153| Build, test, broker quickstart | [README.md](README.md) |
154| Contributing                   | [CONTRIBUTING.md](CONTRIBUTING.md), https://kafka.apache.org/contributing.html |
155| Message protocol / codegen     | [clients/src/main/resources/common/message/README.md](clients/src/main/resources/common/message/README.md) |
156| System tests                   | [tests/README.md](tests/README.md) |
157| JMH benchmarks                 | [jmh-benchmarks/README.md](jmh-benchmarks/README.md) |
158| Client examples                | [examples/README.md](examples/README.md) |
159| Security                       | [SECURITY.md](SECURITY.md) |
160