Codex Plan Validation (Runtime-adaptive) Skill

Codex Plan Validation (Runtime-adaptive)

When to use

complexity: complex
taskType: feature or refactor
context.md exists or was updated

Inputs

analysisContext.* (structured state)
context.md (path: analysisContext.artifacts.contextDocPath)

Runtime Adapter Policy

executionRuntime must be resolved before running this skill.

claude-code: use mcp__codex__codex when available, then Claude fallback if needed.
codex: run the same plan validation rubric directly in the current Codex session (native path), without mcp__codex__codex dependency.

Procedure

Step 1: Resolve Runtime Execution Path (CRITICAL - Do This First)

Determine runtime and select execution path:

If runtime is codex -> skip MCP availability check and use Codex native path.
If runtime is claude-code -> verify Codex MCP availability:

// Try a simple MCP call to check availability
try {
  mcp__codex__codex({
    prompt: "ping",
    sandbox: "read-only",
    cwd: process.cwd()
  })
  // If successful, MCP is available
} catch (error) {
  // MCP not available - proceed with Claude fallback
}

MCP Unavailable Conditions:

Tool not found / not registered
"quota exceeded", "rate limit", "API error", "unavailable"
Connection timeout
Any error response

Step 2-8: Validation Process

Collect the path to context.md (default: {tasksRoot}/{feature-name}/context.md) and read its content
Build delegation prompt using the 7-section format below
If MCP is available (from Step 1):
- Call mcp__codex__codex (include Plan Reviewer instructions in developer-instructions)
- If successful, proceed to step 6
If MCP is unavailable (from Step 1):
- Claude directly performs the plan review following the Plan Reviewer guidelines below
- Add note: "codex-fallback: Claude performed review directly (MCP unavailable)"
- Follow the same MUST DO / MUST NOT DO criteria
If runtime is codex:
- Run plan validation directly in the current Codex session using the same 7-section format and criteria
- Add note: "codex-native: plan validation executed in Codex runtime"
Summarize critical/warning/suggestion items and decide pass/fail
Per docs/public/guidelines/document-memory-policy.md: Store full review in archives/review-v{n}.md, keep only short summary in context.md

Delegation Format

Use the 7-section format:

TASK: Review implementation plan at [context.md path] for completeness and clarity.

EXPECTED OUTCOME: APPROVE/REJECT verdict with specific feedback.

CONTEXT:
- Plan to review: [content of context.md]
- Goals: [what the plan is trying to achieve]
- Constraints: [project constraints]

MUST DO:
- Evaluate all 4 criteria (Clarity, Verifiability, Completeness, Big Picture)
- Reject plans that omit exact files, exact commands, expected fail/pass signals, blocker conditions, review checkpoints, or verification evidence paths
- Simulate actually doing the work to find gaps
- Provide specific improvements if rejecting

MUST NOT DO:
- Rubber-stamp without real analysis
- Provide vague feedback
- Approve plans with critical gaps

OUTPUT FORMAT:
[APPROVE / REJECT]
Justification: [explanation]
Summary: [4-criteria assessment]
[If REJECT: Top 3-5 improvements needed]

Tool Call (Claude Code + MCP Available)

mcp__codex__codex({
  prompt: "[7-section delegation prompt with full context]",
  "developer-instructions": "[contents of plan-reviewer.md]",
  sandbox: "read-only",  // Advisory mode
  cwd: "[current working directory]"
})

Claude Fallback (Claude Code + MCP Unavailable)

When MCP is not available, Claude performs the validation directly:

Apply the same 7-section format as a self-review checklist
Evaluate all 4 criteria:
- Clarity: Are the goals and steps clearly defined?
- Verifiability: Can success be measured objectively with exact commands and evidence paths?
- Completeness: Are exact files, expected signals, blockers, review checkpoints, and evidence paths included?
- Big Picture: Does it align with overall architecture?
Output in the same format: APPROVE/REJECT with justification
Add note indicating fallback mode was used

Codex Native Path (When runtime=codex)

When running in Codex runtime, execute plan validation directly:

Apply the same 7-section format as the validation checklist
Evaluate all 4 criteria (Clarity, Verifiability, Completeness, Big Picture)
- Reject if exact files, commands, fail/pass signals, blocker conditions, review checkpoints, or evidence paths are missing.
Output in the same format: APPROVE/REJECT with justification
Add note: "codex-native: plan validation executed in Codex runtime"

Output (patch)

notes:
  - "codex-plan: [APPROVE/REJECT], warnings=[count]"
  # If fallback was used:
  - "codex-fallback: Claude performed review directly (MCP unavailable)"
  # If Codex runtime native path was used:
  - "codex-native: plan validation executed in Codex runtime"

Agent Skills: Codex Plan Validation (Runtime-adaptive)

Install this agent skill to your local

Skill Files