Truncation Detector & Graceful Terminator MCP Connector for Claude
A+Detects and repairs truncated LLM outputs to restore structural integrity.
This MCP server provides essential diagnostic and recovery tools for handling incomplete Large Language Model (LLM) responses. It identifies if an output was cut off due to token limits by checking for mid-word cutoffs, missing terminal punctuation, or unclosed structural elements like brackets and braces. Using detect_truncation, you can confirm if a response is incomplete. The repair_truncated_text tool uses a deterministic stack-based approach to append the necessary closing characters, ensuring JSON or Markdown structures are syntactically valid. Additionally, validate_structural_integrity allows for verifying the completeness of text in plain_text, json, or markdown formats.
Related Connectors
Output Format Stability Checker MCP
Detect non-deterministic shifts in LLM response structures.
LLM JSON Repair Utility MCP
Fixes malformed, truncated, or dirty JSON strings produced by LLMs.
Output Serializability Checker MCP
Verify if LLM-generated data structures can be safely converted to specific formats without data loss.
LLM JSON Extractor MCP
Isolates valid JSON objects from messy LLM text outputs.