OpenParable
Skip to content
← Back to skills

Commit Conventions

Development
commit-conventions

Commit Conventions is an Agent Skill for development published in the areal-project/AReaL repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

AReaL commit message conventions. MUST load on every git commit -- provides Conventional Commits format with scope inference from file paths.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/areal-project/AReaL/blob/HEAD/.agents/skills/commit-conventions/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/commit-conventions, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Commit Conventions

Commit message conventions and scope inference rules for the AReaL repository.

When to Use

ALWAYS load this skill when making any git commit in AReaL. This includes:

  • Direct commits (git commit)
  • Commits during PR creation ($create-pr / /create-pr)
  • Commits delegated to sub-agents with this skill loaded
  • Any agent workflow that produces a commit

Commit Message Format

<type>(<scope>): <subject>

<body>

[Optional sections:]
Key changes:
- change 1
- change 2

Refs: #123, #456

Type Selection

TypeWhen to Use
featNew feature or capability
fixBug fix
docsDocumentation only
govGovernance or maintainer changes
styleFormatting/style-only changes
refactorCode change without feature/fix
perfPerformance improvement
testAdding or fixing tests
buildBuild system or dependencies
ciCI pipeline or workflow changes
choreBuild, deps, config changes
revertRevert a previous commit

Scope Inference

Infer scope from the primary changed file paths:

File Path PatternScope
areal/workflow/workflow
areal/engine/engine
areal/reward/reward
areal/dataset/dataset
areal/api/api
areal/utils/utils
areal/infra/infra
areal/trainer/trainer
areal/models/models
areal/experimental/experimental
docs/docs
examples/examples
AGENTS.md, .agents/, .claude/, .codex/, .opencode/agents
Multiple areasOmit scope or use broader term

Rules

  • Subject: imperative mood, ~50-72 chars, no trailing period
  • Body: explain "why" not "what", wrap at 72 chars
  • Key changes: bullet list of main modifications (for complex commits with 3+ files)
  • Refs: reference issues/PRs if applicable

Examples

Single file fix:

fix(reward): handle empty completion in gsm8k

Return 0 reward instead of raising exception when
completion string is empty after extraction.

Multi-file feature:

feat(engine): add CPU offload support to ArchonEngine

Enable torch_memory_saver for model offloading during
rollout phase to reduce GPU memory pressure.

Key changes:
- Add offload/onload methods to ArchonEngine
- Integrate with weight update flow
- Handle ROCm compatibility

Docs only:

docs: update algorithm comparison table

Add SAPO and GSPO to the algorithm family documentation
with configuration examples.

Agent/tooling changes:

chore(agents): port review-pr command to OpenCode

Add OpenCode-native commands with task() category
delegation instead of hardcoded model names.

Key changes:
- Create .opencode/command/ with review-pr, create-pr
- Replace hardcoded model routing with platform-native review routing
- Add expert subagent consultation patterns

Governance/maintainer changes:

gov(agents): add maintainer ownership for service modules

Update CODEOWNERS and maintainer references to reflect
current governance responsibilities.

Key changes:
- Add maintainers for agent_service and infra ownership
- Align governance docs with updated reviewer responsibilities