Workspace Builder
Provision another company's logged-in mktg.engineer workspace.
These presets create the same kind of workspace OpenAI is using: company-specific competitors, audiences, experiments, draft PRs, and Profound runs.
Brand Strategy Gate
Tell the agent where the brand wants to win before it drafts experiments.
Coding-agent comparison
PR #1842 adds a Codex workflow review variant, FAQ extraction, Profound gates, middleware traffic split, and a coding-agent evaluation ad test.
No direct strategy match. Keep as monitoring unless Profound risk is severe.Enterprise customer proof
PR #1843 adds a ChatGPT Enterprise proof hub mapped to buyer questions and approved source URLs.
Related to enterprise proof, but needs human approval before PR generation.Enterprise transformation
PR #1844 adds OpenAI enterprise transformation proof and a Mistral challenger ad.
No direct strategy match. Keep as monitoring unless Profound risk is severe.Secure enterprise search and regulated knowledge work
PR #1845 adds ChatGPT Enterprise knowledge-search proof and secure-AI guardrails.
Related to knowledge work, but needs human approval before PR generation.Owned ad specificity
Use OpenAI ad-library creative as controls and generate challengers only when page diffs plus Profound gaps justify spend.
No direct strategy match. Keep as monitoring unless Profound risk is severe.Profound and LLM Mode
Use deterministic mocks for demos, or bring customer keys in production.
Demo mode never stores pasted secrets. Production BYOK should use encrypted Cloudflare secrets or a customer-managed vault and store only credential references in D1.