@mukundakatta

Parse partial / truncated / messy JSON for LLM tool calls and structured outputs.

Validate tool-call args before execution. Returns LLM-friendly retry hints.

Token-budget-aware text handling: count, truncate, and chunk for LLM prompts.

Structured-output enforcer: extract and validate JSON from messy LLM text.

Inject, parse, and strip [N] citation markers in RAG outputs.

Snapshot tests for AI tool-call traces: diff, validate, and catch silent regressions.

Token-aware message truncation: fit a chat history into your model's context budget.

Parse JSON5 (comments, trailing commas, unquoted keys) and emit strict JSON.

IANA timezone conversions with real DST rules.

Network-egress firewall for agent tools: check URLs against a declarative policy.

Character-precise unified diff + patch application + patch parsing.

RAG retrieval IR metrics: recall@k, hit@k, MRR, NDCG@k, evaluate_batch.