One merge standard for everything Codex ships
Codex reads AGENTS.md for durable project guidance and runs across local sessions and cloud tasks in parallel worktrees, under Read Only, Auto, or Full Access. Its own Review panel checks the diff inside the session that wrote it. Baz reviews the Change again independently in GitHub, applies the AGENTS.md rules that match the changed paths, and returns one merge verdict no matter which session or worktree produced the code.
OpenAI Codex tells the agent how to work. Baz keeps the standard.
Native customization tells an agent how to work. Baz turns the durable guidance already stored with your code into scoped requirements it verifies on every Change, and Skills Maintainer opens a pull request when a rule stops matching the code.
Native customization
AGENTS.md, Skills, plugins, MCP, subagents, memories, and task prompts
Where instructions drift
A review inside the session that wrote the code shares its blind spots, and parallel cloud tasks can each read AGENTS.md a little differently.
What Baz adds
Baz discovers AGENTS.md and Skills, then applies the matching requirements during an independent review in GitHub, outside the session that produced the code.
A plan OpenAI Codex can defend
Codex can run read-only while it explores. Baz turns that exploration into a plan grounded in the whole organization, presented for approval before code.
Install the Baz plugin and invoke the plan-with-baz Skill.
Codex runs read-only and searches across repositories through Baz while it explores.
The plan is checked against dependencies and risk before implementation.
Upload the approved plan to Baz for team review, with the relevant repositories, model, and usage recorded.
Result A plan the team reviews and versions in Baz before implementation, linked to the PR that implements it.
One independent merge standard
OpenAI Codex can produce work many ways: Local sessions, cloud tasks, parallel worktrees, Skills, plugins, and subagents. Baz reviews the result after the code exists and decides one thing: Whether the final Change matches team requirements and passes an independent review outside the originating session.
| Baz capability | Evidence used | Decision supported |
|---|---|---|
| AI Coding GuidelinesStandards & Conventions | AGENTS.md, CLAUDE.md, Cursor rules, and Skills matched to changed paths | Did the agent follow the team's durable instructions? |
| Spec ReviewerSandboxed app review | The approved plan, the ticket and designs, and the running application | Did the agent build what the team asked for? |
| Code CorrectnessAI Code Review | Execution paths, API contracts, and consumers in this and other repositories | Does the Change work, and what does it break beyond the edited files? |
| Security ReviewSecurity Review | Advanced Security, SAST-inside signals, and the agent security reviewer | Does the Change introduce an exploitable risk, including in agent code? |
| SessionsSessions docs | What every agent ran, read, found, and skipped | Can the team see how the verdict was reached? |
| MergerMerge Automation | Findings, discussions, approvals, checks, and team merge requirements | Is the Change ready to merge, blocked, or waiting for a human decision? |
Keep every agent’s strengths. Add one control layer.
Each agent keeps its own way to shape behavior. Baz adds the same evidence before code and the same independent verdict after, so the toolchain can change without resetting the standard.
| Capability | Claude Code | OpenAI Codex | Cursor | Devin |
|---|---|---|---|---|
| Durable repo guidance | CLAUDE.md, imported AGENTS.md, .claude/rules | AGENTS.md | Project Rules, AGENTS.md | AGENTS.md |
| Baz before code | Packaged Planner plugin and plan capture | Packaged Planner plugin and plan capture | Baz Rule, MCP, and plan capture | Baz MCP with repo-held guidance |
| Baz after code | Independent review and merge verdict | Independent review and merge verdict | Independent review and merge verdict | Independent review and merge verdict |
| Agent setup reviewed for security | Hooks, Skills, plugins, and MCP servers | Skills, plugins, and MCP servers | Hooks, Skills, plugins, and MCP servers | Repo Skills and MCP servers |
| Guidance kept current | Skills Maintainer PRs to CLAUDE.md and Skills | Skills Maintainer PRs to AGENTS.md and Skills | Skills Maintainer PRs to rules and Skills | Skills Maintainer PRs to AGENTS.md and Skills |
One control layer, whichever agent your team picks
Questions teams ask first
Does Baz review Codex cloud tasks and local sessions the same way?
Yes. Baz reviews the resulting Change with the same standard regardless of whether it came from a local session, a cloud task, or a parallel worktree.
Does Baz replace AGENTS.md?
No. Baz uses the AGENTS.md and Skills your team maintains, normalizes them, and verifies the final Change against the parts that apply to the changed paths.
Can Baz tell which model produced a plan?
Planner records the coding agent and model when the host exposes that information, along with the relevant repositories and final plan.
Let developers keep OpenAI Codex. Keep one standard for the code.
Connect Baz to your repositories and turn the guidance you already maintain into requirements verified on every Change, before code and again before merge.