feat(advisors): enable recursive advisor execution with two new built-in advisors #4622
+2,438
−3
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…-in advisors Add support for recursive (repetitive) advisor execution patterns, allowing advisors to re-invoke themselves or the remaining chain multiple times. This enables advanced patterns like retry logic, iterative refinement, and multi-pass processing. - Add AdvisorUtils.copyChainAfterAdvisor() utility to enable recursive chain invocation - Implement ToolCallAdvisor for recursive tool calling with configurable tool execution - Implement StructuredOutputValidationAdvisor for recursive output validation with retry logic - Add MCP JSON Jackson2 dependency for JSON schema validation - Add test suites for both new advisors and utility methods - Add documentation for recursive advisor patterns Signed-off-by: Christian Tzolov <christian.tzolov@broadcom.com>
…idation feedback loop Refactor retry logic to provide validation error feedback to LLM for self-correction. Extract validation into separate method and augment prompts with error messages on retry. Add comprehensive test coverage for various validation scenarios including nested objects, lists, malformed JSON, and type mismatches. Signed-off-by: Christian Tzolov <christian.tzolov@broadcom.com>
Signed-off-by: Christian Tzolov <christian.tzolov@broadcom.com>
Implements return direct functionality allowing tools to bypass the LLM and return results directly to clients. Adds null safety checks for chatResponse and comprehensive test coverage. Signed-off-by: Christian Tzolov <christian.tzolov@broadcom.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add support for recursive (repetitive) advisor execution patterns, allowing advisors to re-invoke themselves or the remaining chain multiple times. This enables advanced patterns like retry logic, iterative refinement, and multi-pass processing.
Depends on: modelcontextprotocol/java-sdk#617
TODO: The ToolCallAdvisor doesn't implement streaming tool-calling model at the moment. Perhaps we can address this in a follow up issues along with auto-config integrations for the ToolCallAdvisor.