Generate a local Markdown table of contents with duplicate-aware heading anchors.
ReadyRuns in your browser
INPUT & CONTROLS
Prepare your input
Local processing
Extracts ATX and Setext headings, including Markdown headings in lists and quotes. Code and raw HTML blocks are ignored. Labels flatten inline formatting; references decode once outside code and escaped ampersands. Anchors use github-slugger across all headings before level filtering. Other renderers may use different IDs. Missing heading levels do not create empty outline levels. Output is raw Markdown, without a live preview.
Limits: 262,144 input UTF-16 units; 20,000 original lines; 20,000 tokens; token depth 64; 1,000 headings; 4,096 units per normalized label; 300,000 output units. Unsupported controls and malformed surrogates reject. Parsing runs in a terminable worker with a 2,000 ms timeout. Edits, clear, examples and new runs cancel previous work. Web Worker support is required.
Runs locally in a browser worker. Markdown is not sent or stored.
Result
Your result starts here
Review the input and selected options, then use the primary action. Output and any validation messages will appear here.
METHOD & CONTEXT
Understand the result before using it.
How it works
Tokenizes Markdown in a terminable worker, extracts plain heading labels and generates github-slugger anchors before level filtering. The outline uses surviving ancestors only.
Interpretation
Read the result summary first, then inspect the detailed output and any errors or warnings. The result describes the supplied input and selected options. Changing input requires another run.
Assumptions & warnings
Other Markdown renderers may assign different IDs. Empty labels or anchors are skipped. Worker support is required; parsing stops after 2,000 ms. Raw Markdown is shown without rendering.
Inputs are processed in your browser. Follow the input format and size limits shown next to the controls.
Try an example
Select Load example in the input panel, review the selected mode, then use its displayed primary action. Inspect the result before replacing the sample with your own input.