You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When users ask to read, summarize, or analyze non-Markdown sources (PDF, DOCX, PPTX, HTML, image-derived text, or URLs), first convert the source with `/to-markdown`.
2
+
3
+
Guidelines:
4
+
- Prefer `/to-markdown <input> <output.md>` for large inputs.
5
+
- For small inputs, `/to-markdown <input>` and return inline Markdown is acceptable.
6
+
- After conversion, continue analysis using the Markdown output only.
description: Convert files and URLs into Markdown so AI agents can reliably read and reason over mixed-format content.
4
+
origin: opencode-ecc-devcontainer
5
+
---
6
+
7
+
# MarkItDown Converter
8
+
9
+
Use this skill when you need to transform non-Markdown sources (PDF, Office docs, HTML, images with OCR-capable inputs, etc.) into Markdown before analysis.
10
+
11
+
## When to Activate
12
+
13
+
- User asks to summarize or extract data from files that are not already Markdown
14
+
- User provides URLs or documents that should be converted for downstream AI workflows
15
+
- You need a consistent Markdown representation before chunking, indexing, or prompt injection checks
16
+
17
+
## OpenChamber Usage
18
+
19
+
1. Explicit command:
20
+
-`/to-markdown <input> [output.md]`
21
+
2. Natural language:
22
+
- Ask normally (e.g. "このPDFをMarkdown化して"), then route through `/to-markdown`
23
+
24
+
## Notes
25
+
26
+
- If output is large, prefer writing to a file and then summarizing key sections.
27
+
- Preserve the generated Markdown as an artifact when reproducibility is needed.
0 commit comments