---
name: pipeline-assess
description: Codex skill alias for /pipeline-assess. Pre-plan assessment of current codebase and UX state
argument-hint: "[area to assess or feature context]"
---

# Codex Command Alias: /pipeline-assess

<!-- Generated by tools/generate-codex-command-skills.py; edit the source command instead. -->

- This generated skill exposes the Claude slash command `/pipeline-assess` to Codex.
- Treat the user's message that invoked this skill as the command arguments.
- When the embedded command body says `$ARGUMENTS`, substitute those user-supplied arguments.
- Follow the embedded command body as the workflow source of truth.
- Translate Claude-only tool names to Codex equivalents when needed: use `update_plan` for TodoWrite-style ledgers, `request_user_input` for AskUserQuestion-style gates when available, and normal chat questions when that tool is not available.
- If the command body says to launch an agent but no direct subagent tool is available, use the best available Codex delegation tool or execute the referenced agent protocol yourself while clearly noting the degradation.

## Embedded Claude Command

# Pipeline Assess

Standalone pre-plan assessment. Evaluates current codebase state and UX before planning changes.

## Input

<assess_input> #$ARGUMENTS </assess_input>

If the input above is empty, ask: "What area should I assess? Give me a feature area, directory path, or describe what you're planning to change."

## Process

1. Load the assess skill from `plugins/pipeline/skills/assess/SKILL.md`
2. Execute the full assessment protocol (code + UX in parallel)
3. Save the Assessment Brief to `plans/<area-slug>/assessment.html` (HTML + data island, per the assess skill's Artifact Format)
4. Present the brief to the user

## After Assessment

Ask: "Assessment saved to `plans/<slug>/assessment.html` (open it in a browser to review). Want to run the full pipeline from here (`/pipeline`), or use this brief as context for your own planning?"