AI activation
AI Owner Decision
This public mirror gives the owner a launch decision view for OpenAI-assisted WarpIntel analysis. It shows branch labels, owner checklist counts, evidence field counts, activation gate counts, proof targets, and safety booleans while provider keys, model values, prompts, token spend, owner evidence values, protected packets, and private EVE data stay excluded.
Live AI stays gated until protected owner evidence and first-tool smoke are complete.
1 branch blocks live AI activation.
3 item(s) still need protected owner review.
10 evidence field(s) remain protected-review only.
3 criterion item(s) block provider activation.
Owner evidence can be packaged without OpenAI calls, prompts, or token spend.
Every target is no OAuth, no OpenAI call, no token spend, no prompt, and no database write.
Deterministic public tools stay available while OpenAI enhancement remains owner-gated.
No protected admin routes, owner values, provider values, prompts, spend records, or secrets.
AI Owner Decision Brief
Copy, download, or save a browser-local owner packet for the current AI branch recommendation. The packet carries public-safe counts only and does not call OpenAI, submit prompts, spend tokens, call providers, mutate provider state, write database rows, enable live AI, or expose private data.
Copy, download, and browser-local save only.
Keys, prompts, spend records, owner evidence values, and private data stay excluded.
No server-side report, database write, provider call, or account change is created.
Accept owner-reviewed AI activation
Live OpenAI enhancement can only activate after protected owner evidence, first-tool smoke, budget review, and rollback review are complete.
- Public Route
- /ai-readiness
- Requirements
- 6
Repeat AI owner review
Owner can repeat protected review while public pages keep showing counts only and deterministic tools keep working.
- Public Route
- /ai-readiness#owner-evidence-checklist
- Requirements
- 4
Keep OpenAI gated
Recommended until owner evidence is complete; no OpenAI calls, prompt submissions, token spend, provider mutations, or database writes occur.
- Public Route
- /ai-owner-decision
- Requirements
- 3
Activation evidence fields complete
Protected owner review still needs evidence before live AI can activate.
Activation outcome passed
Protected owner review still needs evidence before live AI can activate.
Owner AI activation decision ready
Protected owner review still needs evidence before live AI can activate.
Live AI remains gated by owner review
Public-safe readiness evidence is present.
No provider mutation from evidence capture
Public-safe readiness evidence is present.
Rollback and smoke path ready
Public-safe readiness evidence is present.
AI owner decision page
Shows AI activation branch labels, owner checklist counts, and no-provider-call safety only.
- Route
- /ai-owner-decision
AI owner decision JSON
Returns no-store owner decision branch, checklist, proof target, and safety counts only.
- Route
- /api/ai/owner-decision-readiness
AI readiness page
Shows deterministic fallback, protected-check counts, and owner evidence labels only.
- Route
- /ai-readiness
AI readiness JSON
Returns public-safe AI readiness counts without keys, model values, prompts, or spend records.
- Route
- /api/ai/readiness
Public tools
Keeps analyzer workflows deterministic-first while AI remains owner gated.
- Route
- /tools
Integrations
Shows provider readiness labels without provider account values or secret settings.
- Route
- /integrations
Health JSON
Lets live smoke prove the AI owner decision mirror stayed public-safe.
- Route
- /api/health
Status
Shows aggregate service state without OpenAI keys, prompt text, or token-spend details.
- Route
- /status
Releases
Shows release evidence labels without provider secrets or private EVE data.
- Route
- /releases
Next actions
Shows the owner evidence lane while live AI activation remains gated.
- Route
- /next-actions
No-Secret AI Owner Decision Rule
Public AI owner-decision readiness exposes decision branch labels, recommended branch state, owner checklist counts, owner evidence field counts, activation criteria counts, publish gate counts, proof target outcomes, public links, browser-local brief counts, and no-side-effect booleans only. It does not expose OpenAI keys, model environment values, billing details, raw prompts, pasted analyzer inputs, token-spend records, owner evidence values, provider account data, protected packet bodies, EVE tokens, private EVE data, auth cookies, private logs, or separate-project account details, and it does not start OAuth, call OpenAI, submit prompts, spend tokens, call providers, mutate provider state, write database rows, enable live AI, or expose private data.