Feature Request: Add Token Count Estimation to Output

I'd like to propose adding an estimated token count to the generated output. This would help users know if their generated text fits within their LLM's context window limits.

**Proposed Feature:**
- Add an estimated token count at the beginning of both `llms.txt` and `llms-full.txt` files
- Display format could be something like:
`Estimated Tokens: 12,345`

**Why this would be useful:**
- Helps users immediately know if the generated text will fit their LLM's context window
- Prevents trial-and-error when loading large text files into LLMs
- Makes it easier to split content into appropriate chunk sizes if needed

**Implementation Suggestions:**
- Could use libraries like `tiktoken` or a simple character-based approximation
- Token count could be placed in a header section or metadata block at the start of the file