Expand Codex and Pi workflow coverage

.github/workflows/ab-testing-advisor.md

@@ -33,10 +33,10 @@ safe-outputs:
     title-prefix: "[ab-advisor] "
 description: Daily A/B testing advisor that picks a random agentic workflow without an experiments section, devises an experiment campaign to improve it, and creates a GitHub issue with the implementation task
 emoji: 🧪
-model: copilot/gpt-5.4
+model: openai/gpt-5.4
 engine:
-  bare: true
   id: pi
+  model-provider: openai
 strict: true
 timeout-minutes: 30
 tools:

.github/workflows/ace-editor.md

@@ -10,6 +10,10 @@ on:
     events: [pull_request_comment]
 max-daily-ai-credits: 10000
 timeout-minutes: 10
+engine:
+  id: codex
+  model-provider: openai
+model: openai/gpt-5.4
 strict: false
 permissions:
   pull-requests: read

.github/workflows/agent-persona-explorer.md

@@ -4,9 +4,10 @@ emoji: "🎭"
 description: Explores agentic-workflows custom agent behavior by generating software personas and analyzing responses to common automation tasks
 on: daily
 max-daily-ai-credits: 10000
-model: copilot/gpt-5.4
+model: openai/gpt-5.4
 engine:
   id: pi
+  model-provider: openai
 permissions:
   contents: read
   actions: read

.github/workflows/agentic-token-trend-audit.md

@@ -98,6 +98,10 @@ evals:
   - id: trend_report_created
     question: Was a trend report created with per-workflow AIC data for the requested period?
 
+engine:
+  id: codex
+  model-provider: openai
+model: openai/gpt-5.4
 ---
 
 # On-Demand Agentic Workflow AIC Trend Audit

.github/workflows/api-consumption-report.md

@@ -11,11 +11,11 @@ permissions:
   issues: read
   pull-requests: read
   discussions: read
+  copilot-requests: write
 tracker-id: api-consumption-report-daily
 engine:
-  id: claude
-  mcp:
-    tool-timeout: 10m
+  id: codex
+  model-provider: github
 sandbox:
   agent:
     runtime: cloud-hypervisor
@@ -51,6 +51,7 @@ evals:
   - id: report_with_charts_created
     question: Was a report or discussion created with trending charts and quota analysis?
 
+model: copilot/mai-code-1-flash-picker
 ---
 
 # GitHub API Consumption Report Agent

.github/workflows/artifacts-summary.md

@@ -11,8 +11,9 @@ permissions:
   actions: read
   copilot-requests: write
 engine:
-  id: codex
-model: copilot/mai-code-1-flash-picker
+  id: pi
+  model-provider: openai
+model: openai/gpt-5.4
 network:
   allowed:
     - defaults

.github/workflows/audit-workflows.md

@@ -28,9 +28,8 @@ experiments:
     start_date: "2026-07-03"
     issue: 43177
 engine:
-  id: claude
-  mcp:
-    tool-timeout: 10m
+  id: codex
+  model-provider: openai
 tools:
   cli-proxy: true
   agentic-workflows:
@@ -68,6 +67,7 @@ evals:
   - id: issues_identified_or_noop
     question: Were issues, missing tools, errors, and improvement opportunities identified, or was noop used when no problems were found?
 
+model: openai/gpt-5.4
 ---
 
 # Agentic Workflow Audit Agent

.github/workflows/auto-triage-issues.md

@@ -16,9 +16,10 @@ permissions:
   contents: read
   issues: read
   copilot-requests: write
-model: copilot/gpt-5.4
+model: openai/gpt-5.4
 engine:
   id: pi
+  model-provider: openai
 strict: true
 network:
   allowed:

.github/workflows/avenger.md

@@ -18,9 +18,10 @@ env:
   GH_AW_HARNESS_MAX_RETRIES: "4"
 tracker-id: avenger-ci
 max-turns: 50
-model: claude-haiku-4-5
+model: openai/gpt-5.4
 engine:
-  id: claude
+  id: codex
+  model-provider: openai
 network:
   allowed:
     - defaults