feat(Designer): Add reasoning flag to agent loop in consumption

[Elaina-Lee]

Commit Type

• feature - New functionality
• fix - Bug fix
• refactor - Code restructuring without behavior change
• perf - Performance improvement
• docs - Documentation update
• test - Test-related changes
• chore - Maintenance/tooling

Risk Level

• Low - Minor changes, limited scope
• Medium - Moderate changes, some user impact
• High - Major changes, significant user/system impact

What & Why

Adding resoning flag to the agent loop

Impact of Change

• Users: Users can now find and add reasoning flag as a advanced parameter option
• Developers: none
• System: none

Test Plan

• Unit tests added/updated
• E2E tests added/updated
• Manual testing completed
• Tested in: Local standalone

Contributors

Screenshots/Videos

[github-actions]

🤖 AI PR Validation Report

PR Review Results

Thank you for your submission! Here's detailed feedback on your PR title and body compliance:

✅ PR Title

• Current: feat(Designer): Add reasoning flag to agent loop in consumption
• Issue: None significant. The title follows the conventional format (type(scope): summary) and is descriptive.
• Recommendation: Keep as-is. (Optional: you can make it slightly more explicit about the field name if you prefer, e.g. feat(Designer): Add reasoningEffort dropdown to agent loop (consumption))

---

✅ Commit Type

• Properly selected (feature)
• Only one selection is checked; this is correct.

---

✅ Risk Level

• Selected Low and the PR has a risk:low label. The code diff is a small manifest addition (single file, +29 lines) that adds a UI dropdown option — this is consistent with a low-risk change.

---

⚠️ What & Why

• Current: Adding resoning flag to the agent loop
• Issue: Short and contains a typo (resoning → reasoning). It is too brief to be fully informative (doesn't name the manifest file or the exact field added, default behavior, or motivation beyond a one-line summary).
• Recommendation: Expand to 2–3 sentences and fix the typo. Example you can paste into the PR body:
  • Add reasoningEffort dropdown to the agent loop manifest (libs/logic-apps-shared/src/designer-client-services/lib/consumption/manifests/agentloop.ts). This adds an advanced parameter "reasoningEffort" (options: low, medium, high, none, minimal) so designers can control reasoning effort for agent loop actions. This change is needed to expose the reasoning effort option in the designer UI for consumption flows.
  • Also mention default behavior (if any) and whether existing saved workflows will be impacted.

---

⚠️ Impact of Change

• Issue: The Impact section is present but minimal (Developers: none, System: none). Even small manifest changes can affect developers or pipeline consumers if they read or validate manifests; please be explicit.
• Recommendation: Clarify impacts. Suggested wording:
  • Users: Designers can now select a reasoning effort for agent loops in the designer UI; no breaking user workflow changes expected by default.
  • Developers: Manifest schema updated — if there are any code paths that read/validate this manifest, they may need to accept the new reasoningEffort field. If none do, say so explicitly.
  • System: No runtime behavior changes outside of the UI. If the value is forwarded to back-end services or telemetry, list that.

---

⚠️ Test Plan

• Assessment: The PR marks manual testing completed and notes Tested in: Local standalone. The code diff shows no unit or E2E tests were added.

• Action / Recommendation: Either add an automated test or expand the Test Plan with detailed manual test steps and rationale for why automated tests are not required. Example manual test steps to include:

  1. Open the designer in Local standalone build.
  2. Add or open an agent loop action and open Advanced parameters.
  3. Verify the reasoningEffort dropdown appears with options: Low, Medium, High, None, Minimal.
  4. Select each option and save; confirm manifest contains the corresponding reasoningEffort value.

  If you can add tests, consider:

  • A unit test validating the manifest schema contains the reasoningEffort property, or
  • An E2E/UI test that asserts the dropdown is visible and persists the selection.

  If automated tests are not feasible for a manifest-only UI change, add the explicit manual test steps to the PR body and explain why automated tests are deferred.

---

⚠️ Contributors

• Assessment: Blank.
• Recommendation: This section is optional, but it's good practice to credit reviewers, PMs, or designers who helped. Add any names or GitHub mentions if applicable; otherwise you can leave blank.

---

✅ Screenshots/Videos

• Assessment: Screenshot included in PR body. Good — helps reviewers validate UI changes.

---

Summary Table

Section	Status	Recommendation
Title	✅	Keep as-is (optional small wording tweak)
Commit Type	✅	None
Risk Level	✅	Matches code diff; low risk
What & Why	⚠️	Fix typo and expand to include file name, field name, default, and rationale
Impact of Change	⚠️	Add more specific impacts for Users/Developers/System
Test Plan	⚠️	Add manual test steps or an automated test; explain why tests are omitted if so
Contributors	⚠️	Optional: credit contributors if any
Screenshots/Videos	✅	Good — screenshot provided

---

Final message
This PR largely passes the PR title/body checks. The advised risk level is Low (matches the PR label risk:low and the small manifest-only change in the diff). Please update the PR body to:

• Fix the typo in What & Why (resoning → reasoning).
• Expand What & Why with the file path and exact field added and why the change is needed.
• Expand Impact of Change with clear notes for Developers and confirm whether this is backward-compatible.
• Add explicit manual test steps to the Test Plan or include an automated test verifying the manifest/UI behavior (recommended).

After those small updates, this PR will be fully documented and ready from a PR title/body perspective. Thank you for the clear screenshot and for keeping the risk level accurate.

---

Last updated: Fri, 06 Feb 2026 17:48:54 GMT

[github-actions]

📊 Coverage check completed. See workflow run for details.

[Copilot]

Pull request overview

This PR adds the reasoningEffort parameter as an advanced option to the agent loop configuration in the consumption manifest. This allows users to control the level of reasoning effort applied by AI models during agent loop execution.

Changes:

• Added reasoningEffort field to agentChatCompletionSettings in the consumption agent loop manifest
• Configured as a dropdown with five options: low, medium, high, none, and minimal
• Follows the same implementation pattern as the standard manifest