diverge-codexlisted
Install: claude install-skill scdenney/open-science-skills
# Diverge with a fresh Codex context
Use a fresh subagent to reduce anchoring from the lead agent's conversation history. Same brainstorm-then-select discipline as `$diverge`; both contexts run Codex, so this buys a clean context rather than cross-model diversity — never describe it as an independent model family.
Read [`../diverge/references/creative-preference-optimization.md`](../diverge/references/creative-preference-optimization.md) only when the user asks for the rationale or when refining the creativity criteria.
## Gate delegation
Delegate only when the user explicitly invokes `$diverge-codex` or asks for a subagent, independent context, or delegated brainstorm. If the skill was loaded implicitly without such a request, run the ordinary `$diverge` workflow locally.
## Phase 1: brief the brainstormer
Extract the task, hard constraints, existing artifacts, and acceptance criteria. Spawn one fresh Codex subagent with only the task-local context it needs, withholding your own preferred solution or likely implementation. Give it this contract:
```text
Generate 3–5 conceptually distinct approaches to TASK. Do not implement.
Label every approach [Novel], [Surprising], [Diverse], or [Conventional].
Include at least one [Novel] and one [Surprising] approach. For each, provide:
1. Core mechanism — one sentence.
2. How it works and why it differs — two or three sentences.
3. Main tradeoff — one sentence.
Do not restate one idea with different vocabulary. Return only the