| name | chatgpt-share-markdown |
| description | Convert public ChatGPT share links into clean Markdown transcripts for agent context. Use when a user provides a https://chatgpt.com/share/... URL and wants the conversation summarized, archived, or reused as input context. |
Chatgpt Share Markdown
Convert shared ChatGPT conversations into role-labeled Markdown using ./scripts/chatgpt-share-markdown.cs.
dotnet run --file ./scripts/chatgpt-share-markdown.cs -- "https://chatgpt.com/share/<id>"
Workflow
- Run the script with a public
chatgpt.com/share/... URL.
- Inspect the generated Markdown for media placeholders and formatting.
- Use the Markdown directly as context, or trim it to only the relevant turns.
Commands
Export to stdout:
dotnet run --file ./scripts/chatgpt-share-markdown.cs -- "https://chatgpt.com/share/<id>"
Export to a file:
dotnet run --file ./scripts/chatgpt-share-markdown.cs -- "https://chatgpt.com/share/<id>" --output "/tmp/chat-context.md"
Output
- Conversation title as H1
- Metadata block with source URL and optional update timestamp/model
- One section per turn:
## User and ## Assistant by default (## Tool optional)
- Markdown-friendly content with fenced code blocks and basic multimodal placeholders
- Internal tool-action payloads and reasoning recaps are omitted by default to keep context clean
Optional flags:
--include-tool-turns --include-internal-actions --include-reasoning
Troubleshooting
- If the URL is not a public share link, request
https://chatgpt.com/share/....
- If parsing fails, rerun once; share page payload shape can change.
- If output looks incomplete, keep the parser output and supplement manually only for missing turns.
