shardingsphere AGENTS.md

ShardingSphere Codex Development Guide

This repository guide is written for Codex. Keep only stable, repository-wide rules here and rely on Codex for ordinary coding competence. Follow every explicit rule literally; do not replace a repository rule with general judgment unless that rule authorizes it. Paths are relative to the repository root.

Instruction Sources and Routing

1. CODE_OF_CONDUCT.md is the authority for contribution, Java, and unit-test style. Inspect the applicable section before changing code or tests and record it when it controls a decision.
2. Before Maven, E2E, Proxy startup, database clients, IDE/MCP run configurations, commands likely to output more than 100 lines, or large structured analysis, read or reuse .codex/context/token-efficiency.md and follow its Mandatory Execution Contract.
3. Before the first code-affecting write in a task, read .codex/skills/code-implementation/SKILL.md through EOF and follow it. Its implementation reference defines the repository Codex design style, and every in-scope violation in the effective candidate is a required finding. Code-affecting writes include every production, test, script, or other implementation artifact, including build logic, generated source, and behavior-affecting configuration. Read-only analysis or review, diagnosis-only work, and prose-only changes do not activate this write workflow. Reading a Skill never grants authority or expands scope.
4. Automatic Skill catalog discovery is diagnostic only. If the catalog does not expose code-implementation but the exact repository file is readable, load it by exact path and continue. If a canonical source required for the current task is missing or unreadable, keep ordinary target writes read-only and report the blocker. Only an exact user-authorized policy or harness repair may use the constrained recovery procedure in .codex/harness/agents/policy-maintenance.md; that exception never applies to production or test code.
5. For read-only code/test analysis, planning, design, or review, read .codex/skills/code-implementation/references/rules/implementation.md, .codex/skills/code-implementation/references/rules/testing.md, and .codex/skills/code-implementation/references/rules/artifact-removal-and-contract-impact.md through EOF without activating the write workflow. Apply each contract, impact, or removal rule only if triggered; read .codex/skills/code-implementation/references/rules/non-regression.md through EOF when reviewing a current-task candidate.
6. Before every analysis or change that classifies an artifact as unused or removable, read .codex/skills/code-implementation/references/rules/artifact-removal-and-contract-impact.md through EOF and apply its complete evidence gate. When a current-task edit removes the last production consumer and the audit proves that no compatibility contract remains, apply its single-model convergence rule. This read-only route does not activate the code-writing workflow.
7. For runtime diagnosis, read .codex/context/runtime-triage.md through EOF. Diagnosis remains read-only; if the user authorizes a fix, activate code-implementation before the first code-affecting write.
8. Before selecting optional cross-cutting Skills for external-version decisions, public-contract design, unexpected failures, performance work, simplification, threat modeling, or high-risk decisions, read .codex/context/cross-cutting-skills.md through EOF and apply its exact triggers and limits.
9. Before choosing, running, or assessing repository verification in any task, read .codex/skills/code-implementation/references/verification.md through EOF. Before handoff, read .codex/context/change-completion.md through EOF and apply every applicable completion step.
10. Use repository Skills for specialized workflows instead of reproducing their detailed instructions here. Keep task-specific notes in the task, issue, or PR; do not add session notes to this file.