/find-issue

The find_issues endpoint is an intelligent diff analyzer that identifies potential breaking changes and their ripple effects across your codebase. Think of it as having a vigilant code reviewer who instantly understands how your changes impact every connected component, service, and dependency.

What Makes Deep Issues Unique

Unlike traditional diff tools that show what changed, Deep Issues:

  • Predicts impact across your entire codebase, not just the changed files

  • Identifies breaking changes before they reach production

  • Traces dependency chains to find affected consumers

  • Understands semantic changes beyond syntax modifications


API/MCP Reference

Request Format

{
  "tool": "issue",
  "parameters": {
    "input": [
      "@@ -10,7 +10,7 @@ router = APIRouter()",
      "",
      "[email protected](\"/foo\")",        // Original endpoint
      "[email protected](\"/foo/bar\")",    // Modified endpoint - potential breaking change
      " async def get_users()"
    ],
    "repositories": ["backend/api", "frontend/app"],  // Check impact across repos
    "session_id": "analysis-123"                      // Track conversation context
  }
}

Parameters

Parameter
Type
Required
Description

input

string

Yes

Git diff format changes or textual description of modifications. Include context lines for better analysis.

repositories

array

No

List of repository identifiers to analyze for impact. Format: org/repo

session_id

string

No

Unique identifier to maintain context across multiple analyses in a session.

  • Include sufficient context - Provide several lines before and after changes for accurate analysis

  • Specify affected repositories - Include all repos that might consume the changed code

  • Group related changes - Analyze related modifications together for comprehensive impact assessment

  • Use session_id for complex migrations - Chain multiple analyses for large refactoring projects

  • Include semantic descriptions - Add comments about the intent of changes for better recommendations

Last updated

Was this helpful?