Cli Structured Output + Codex Collector

✓ Guardrail ↓ Collector Experimental Devex Build And Ci

Enforce Cli Structured Output using data collected by Codex Collector. Automatically check devex build and ci standards on every PR.

Guardrail: Ensures Codex CLI invocations in CI use structured JSON output instead of plain text. JSON output makes AI automation deterministic and parseable.

Data Source: Detects OpenAI Codex CLI invocations in CI pipelines. Records command strings, versions, and flags for policy-level analysis. Writes to ai.native.codex under the unified ai.* namespace.

Get Started View on GitHub

How Codex Collector Powers This Guardrail

The Codex Collector gathers metadata from your systems. This data flows into Lunar's Component JSON, where the Cli Structured Output guardrail evaluates it against your standards.

When enabled, this check runs automatically on every PR and in AI coding workflows, providing real-time enforcement with actionable feedback.

↓ Codex Collector Gathers Data Collector

Extracts metadata from code, configs, and tool outputs

{ } Component JSON

Data centralized in structured format for evaluation

✓ Cli Structured Output Checks Guardrail

Pass/fail result with actionable feedback in PRs

Quick Start Configuration

Add both the collector and policy to your lunar-config.yml to enable this guardrail.

📄 lunar-config.yml

# Step 1: Enable the Codex Collector
collectors:
  - uses: github://earthly/lunar-lib/collectors/codex@v1.0.5
    # with: ...

# Step 2: Enable the Codex Guardrails
policies:
  - uses: github://earthly/lunar-lib/policies/codex@v1.0.5
    include: [cli-structured-output]
    # with: ...

What Codex Collector Collects

This collector gathers the following data that the Cli Structured Output guardrail evaluates.

Collector code

`instruction-file`

Discovers CODEX.md instruction files across the repository. Records file path, line count, byte size, and whether it is a symlink to AGENTS.md. Writes to ai.native.codex.instruction_file and appends to ai.instructions.all[] for normalized cross-tool access.

Collector ci-after-command

`cicd`

Detects OpenAI Codex CLI invocations in CI pipelines. Records the command string and version for policy-level flag analysis (dangerous flags, sandbox mode, approval mode).

Example Data Flow

Here's an example of the data that Codex Collector writes to the Component JSON, which Cli Structured Output then evaluates.

{ } component.json From Codex Collector

{
  "ai": {
    "instructions": {
      "all": [
        {
          "path": "CODEX.md",
          "dir": ".",
          "filename": "CODEX.md",
          "lines": 60,
          "bytes": 2400,
          "is_symlink": false,
          "symlink_target": null
        }
      ]
    },
    "native": {
      "codex": {
        "instruction_file": {
          "exists": true,
          "path": "CODEX.md",
          "lines": 60,
          "bytes": 2400,
          "is_symlink": false,
          "symlink_target": null
        },
        "cicd": {
          "cmds": [
            {
              "cmd": "codex exec --json 'review this PR'",
              "tool": "codex",
              "version": "0.1.0"
            }
          ]
        }
      }
    }
  }
}

Configuration Options

Codex Guardrails Inputs

Input	Required	Default	Description
`dangerous_flags`	Optional	`--dangerously-bypass-approvals-and-sandbox,--yolo,--full-auto`	Comma-separated dangerous flags for Codex CLI

Learn More

Guardrail · Codex Guardrails Cli Structured Output View all compatible integrations

→

Collector Codex Collector View full documentation

→

Ready to Automate Your Standards?

See how Lunar can turn your AGENTS.md, engineering wiki, compliance docs, or postmortem action items into automated guardrails with our 200+ built-in guardrails.

Works with any process

AI agent rules & prompt files

Post-mortem action items

Security & compliance policies

Testing & quality requirements

Automate Now

Paste your AGENTS.md or manual process doc and get guardrails in minutes

Book a Demo