LCORE-1838: Wire MCP require approval

[asimurka]

Description

Wires configured MCP approval policy to MCP creation process. Distinguishes between simple (string) and complex policies.

Type of change

• Refactor
• New feature
• Bug fix
• CVE fix
• Optimization
• Documentation Update
• Configuration Update
• Bump-up service version
• Bump-up dependent library
• Bump-up library or tool used for development (does not change the final image)
• CI configuration change
• Konflux configuration change
• Unit tests improvement
• Integration tests improvement
• End to end tests improvement
• Benchmarks improvement

Tools used to create PR

Identify any AI code assistants used in this PR (for transparency and review context)

• Assisted-by: Cursor

Related Tickets & Documents

• Related Issue #
• Closes # LCORE-1838

Checklist before requesting a review

• I have performed a self-review of my code.
• PR has passed all pre-merge test jobs.
• If it is a core feature, I have added thorough tests.

Testing

• Please provide detailed steps to perform tests related to this code change.
• How were the fix/results from this change verified? Please provide relevant screenshots or results.

Summary by CodeRabbit

• Bug Fixes

  • Fixed approval-filter handling so generated tools respect server approval settings (supports legacy and structured approval configs), ensuring correct require_approval behavior.

• Tests

  • Added unit tests validating default and configured approval behaviors and structured approval translations.

[asimurka]

Needs rebase on #1773

[coderabbitai]

Warning

Review limit reached

@asimurka, we couldn't start this review because you've used your available PR reviews for now.

Your plan includes 1 review of capacity. Refill in 42 minutes and 25 seconds.

Your organization has run out of usage credits. Purchase more in the billing tab.

⌛ How to resolve this issue?

After more review capacity refills, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than trial, open-source, and free plans. In all cases, review capacity refills continuously over time.

Please see our FAQ for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: 426ff414-b7e0-48cc-a8b8-4b2ec2414fc5

📥 Commits

Reviewing files that changed from the base of the PR and between 11cd6a0 and 0d1165c.

📒 Files selected for processing (3)

• src/app/endpoints/responses_telemetry.py
• src/utils/responses.py
• tests/unit/utils/test_responses.py

Walkthrough

The PR updates MCP tool construction in Responses to derive require_approval from ModelContextProtocolServer.require_approval (supporting string and structured ApprovalFilter forms), adds tests for the behavior, and fixes a telemetry import path.

Changes

MCP Approval Filter Configuration Propagation

Layer / File(s)	Summary
Approval Filter Integration in MCP Tool Construction src/utils/responses.py	Import ApprovalFilter and change MCP tool construction to read require_approval from mcp_server config, supporting both string form ("always"/"never") and structured ApprovalFilter(always=..., never=...) form.
Unit Tests for require_approval Handling tests/unit/utils/test_responses.py	Add aliased Llama Stack ApprovalFilter import and local config ApprovalFilter import; extend TestGetMCPTools with assertions for default "never" and new async tests validating propagation of string and filter-based require_approval configs.
Telemetry import path update src/app/endpoints/responses_telemetry.py	Change ResponsesContext import to come from models.common.responses.contexts instead of models.common.responses.responses_context.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• lightspeed-core/lightspeed-stack#1773: Introduces the require_approval and ApprovalFilter configuration models that this PR propagates into MCP tool definitions.

Suggested reviewers

• tisnik
• max-svistunov
• jrobertboos

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'LCORE-1838: Wire MCP require approval' directly and clearly describes the main change: wiring the MCP require_approval configuration into the MCP tool creation process.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

✨ Simplify code

• Create PR with simplified code

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[asimurka]

Set up by the model automatically

[jrobertboos]

LGTM

[tisnik]

LGTM

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@src/utils/responses.py`:
- Around line 736-748: The explicit MCP tool path doesn’t propagate the server’s
require_approval mapping, so tools passed via
apply_mcp_headers_to_explicit_tools can bypass the ApprovalFilter; update
apply_mcp_headers_to_explicit_tools to copy/convert mcp_server.require_approval
into the tool (same mapping logic used where get_mcp_tools constructs
InputToolMCP: if require_approval is a string keep it, else build
ApprovalFilter(always=..., never=...)) and ensure
InputToolMCP.server_label/server_url are paired with the require_approval value;
add a regression test exercising resolve_tool_choice (or
resolve_client_tool_choice) that supplies an explicit tool from an MCP server
configured with "always" or an ApprovalFilter and asserts the resulting tool has
the require_approval policy applied.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: f43f80ee-04e5-4ca1-bbd3-74f9fcdd1ce5

📥 Commits

Reviewing files that changed from the base of the PR and between 542be8b and 11cd6a0.

📒 Files selected for processing (3)

• src/app/endpoints/responses_telemetry.py
• src/utils/responses.py
• tests/unit/utils/test_responses.py

📜 Review details

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (10)

• GitHub Check: spectral
• GitHub Check: unit_tests (3.13)
• GitHub Check: unit_tests (3.12)
• GitHub Check: E2E: library mode / ci / group 1
• GitHub Check: E2E: server mode / ci / group 2
• GitHub Check: E2E: library mode / ci / group 2
• GitHub Check: E2E: library mode / ci / group 3
• GitHub Check: E2E: server mode / ci / group 1
• GitHub Check: Konflux kflux-prd-rh02 / lightspeed-stack-on-pull-request
• GitHub Check: E2E Tests for Lightspeed Evaluation job

🧰 Additional context used

📓 Path-based instructions (3)

src/**/*.py

📄 CodeRabbit inference engine (AGENTS.md)

src/**/*.py: Use absolute imports for internal modules: from authentication import get_auth_dependency
Llama Stack imports: Use from llama_stack_client import AsyncLlamaStackClient
Check constants.py for shared constants before defining new ones
All modules must start with descriptive docstrings explaining purpose
Use logger = get_logger(__name__) from log.py for module logging
All functions must have complete type annotations for parameters and return types, use modern syntax (str | int), and include descriptive docstrings
Use snake_case with descriptive, action-oriented names for functions (get_, validate_, check_)
Avoid in-place parameter modification anti-patterns; return new data structures instead of modifying function parameters
Use async def for I/O operations and external API calls
Use standard log levels with clear purposes: debug() for diagnostic info, info() for program execution, warning() for unexpected events, error() for serious problems
All classes must have descriptive docstrings explaining purpose and use PascalCase with standard suffixes: Configuration, Error/Exception, Resolver, Interface
Abstract classes must use ABC with @abstractmethod decorators
Follow Google Python docstring conventions with required sections: Parameters, Returns, Raises, and Attributes for classes

Files:

• src/app/endpoints/responses_telemetry.py
• src/utils/responses.py

src/app/**/*.py

📄 CodeRabbit inference engine (AGENTS.md)

src/app/**/*.py: FastAPI dependencies: Import from fastapi module for APIRouter, HTTPException, Request, status, Depends
Use FastAPI HTTPException with appropriate status codes for API endpoints and handle APIConnectionError from Llama Stack

Files:

• src/app/endpoints/responses_telemetry.py

tests/**/*.py

📄 CodeRabbit inference engine (AGENTS.md)

tests/**/*.py: Use pytest for all unit and integration tests; do not use unittest
Use pytest.mark.asyncio marker for async tests

Files:

• tests/unit/utils/test_responses.py

🧠 Learnings (2)

📚 Learning: 2026-04-06T20:18:07.852Z

Learnt from: major
Repo: lightspeed-core/lightspeed-stack PR: 1463
File: src/app/endpoints/rlsapi_v1.py:266-271
Timestamp: 2026-04-06T20:18:07.852Z
Learning: In the lightspeed-stack codebase, within `src/app/endpoints/` inference/MCP endpoints, treat `tools: Optional[list[Any]]` in MCP tool definitions as an intentional, consistent typing pattern (used across `query`, `responses`, `streaming_query`, `rlsapi_v1`). Do not raise or suggest this as a typing issue during code review; changing it in isolation could break endpoint typing consistency across the codebase.

Applied to files:

• src/app/endpoints/responses_telemetry.py

📚 Learning: 2026-02-23T14:56:59.186Z

Learnt from: asimurka
Repo: lightspeed-core/lightspeed-stack PR: 1198
File: src/utils/responses.py:184-192
Timestamp: 2026-02-23T14:56:59.186Z
Learning: In the lightspeed-stack codebase (lightspeed-core/lightspeed-stack), do not enforce de-duplication of duplicate client.models.list() calls in model selection flows (e.g., in src/utils/responses.py prepare_responses_params). These calls are considered relatively cheap and removing duplicates could add unnecessary complexity to the flow. Apply this guideline specifically to this file/context unless similar performance characteristics and design decisions are documented elsewhere.

Applied to files:

• src/utils/responses.py

🔇 Additional comments (1)

src/app/endpoints/responses_telemetry.py (1)

14-14: Approve ResponsesContext import path update in src/app/endpoints/responses_telemetry.py

• Search found no remaining from models.common.responses.responses_context import ... / import * references, so the rename to models.common.responses.contexts appears consistently applied.