fix: Dev to main PR

[Dhruvkumar-Microsoft]

Purpose

This pull request adds and enforces a strict policy to suppress citation markers and attribution in agent responses, particularly for agents using a knowledge base. It also updates tests to ensure these new instructions are correctly applied. The main changes are grouped below:

Knowledge Base Citation Suppression:

• Added a _KNOWLEDGE_BASE_NO_CITATIONS_PROMPT string to agent_factory.py that instructs agents not to include any citation markers, source-reference tokens, attribution markers, or footnotes in their responses.
• Updated create_agent_from_config in agent_factory.py to append the no-citations prompt to agent instructions when a knowledge base is configured.

Prompt and Helper Updates:

• Added the no-citations instruction to the final answer prompt in get_magentic_prompt_kwargs to ensure plan review summaries also suppress citations.

Testing Enhancements:

• Mocked KnowledgeBaseConfig in test setup and ensured agent objects support use_knowledge_base and knowledge_base_name fields. [1] [2] [3]
• Added and extended tests in test_agent_factory.py to verify that knowledge-base-backed agents receive the new citation suppression instructions. [1] [2]
• Added a test in test_plan_review_helpers.py to confirm that final prompts suppress citations as intended.

Does this introduce a breaking change?

• Yes
• No

How to Test

• Get the code

git clone [repo-address]
cd [repo-name]
git checkout [branch-name]
npm install

• Test the code

What to Check

Verify that the following are valid

• ...

Other Information

[github-actions]

Coverage Report •

File	Stmts	Miss	Cover	Missing
agents
agent_factory.py	75	1	98%	148
orchestration
plan_review_helpers.py	129	32	75%	306, 310, 312–313, 319, 323–326, 328–329, 332, 334–341, 343–344, 346–347, 349–354, 527–528
TOTAL	3835	552	85%

Tests	Skipped	Failures	Errors	Time
837	0 💤	0 ❌	0 🔥	5.987s ⏱️

[Copilot]

Pull request overview

This pull request enforces a “no citations / no attribution markers” response policy for knowledge-base-backed agents and for the orchestrator’s final answer prompt, and updates unit tests to validate the new prompt behavior.

Changes:

• Added a KB-specific “RESPONSE CITATION POLICY” prompt segment and appended it to agent instructions when a knowledge base is configured.
• Updated the orchestrator final-answer prompt rules to suppress citation/attribution markers.
• Expanded unit tests to cover KB config wiring and to assert the citation-suppression policy is present in generated prompts.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
src/backend/agents/agent_factory.py	Adds and conditionally appends a KB-specific no-citations prompt to agent instructions when kb_config is present.
src/backend/orchestration/plan_review_helpers.py	Extends final answer rules to suppress citation markers/attribution in orchestrator responses.
src/tests/backend/agents/test_agent_factory.py	Updates module stubs and adds a test asserting KB-backed agents receive the no-citations instructions.
src/tests/backend/orchestration/test_plan_review_helpers.py	Adds a test asserting the final prompt includes the no-citations instruction.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[Copilot]

Pull request overview

Copilot reviewed 25 out of 26 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 26 out of 27 changed files in this pull request and generated 2 comments.

Suppressed comments (1)

Previously missed (1) — in code that hasn't changed since the last review.

src/tests/backend/orchestration/test_plan_review_helpers.py:240

• Test name is misleading: this test doesn’t vary the team, it just verifies the default final_answer_prompt contains the citation-suppression rule when has_user_responses=False. Renaming it to reflect the actual condition would make the suite easier to scan and keep consistent with nearby test names.

    def test_given_any_team_when_called_then_final_suppresses_citations(self):