Who this is for
Developers who form a clear opinion reading a PR but lose half of it typing comment by comment.
The moment this saves you
I read a PR carefully, form a clear opinion, and then lose half my feedback typing it out comment by comment, so the review ends up thinner than the thought I actually had.
See it work
Messy spoken thought in. A clean, structured artifact out.
Reviewing Sam's PR on the export feature. The big blocker is the export runs synchronously in the request handler, with a large dataset that'll time out, this needs to move to a background job before we merge. Couple of suggestions, the error handling swallows exceptions silently in the parse function, we should at least log them, and the function names like doStuff aren't great, could be clearer. Nit, there's a commented-out block on line 40 that should go. On the plus side, the test coverage is genuinely great, they covered the edge cases I'd have missed. Overall good work, just needs the async fix.
Review, Sam's export PR
Blockers (must fix before merge)
- Export runs synchronously in the request handler, will time out on large datasets. Move to a background job.
Suggestions
parseswallows exceptions silently, at least log them.- Unclear function names (e.g.
doStuff), make them descriptive.
Nits
- Commented-out block on line 40 should be removed.
Praise
- Test coverage is genuinely great, edge cases well handled.
Overall: good work, needs the async fix.
The workflow
Record a voice note
Hit the hotkey and talk, no formatting, no typing.
Tag it with this context
Contextli shapes your words into the structured output above.
Find it later
Everything's searchable and organised by context.
Pull it into Claude or ChatGPT
Bring your contexts straight into your AI tools with the Contextli MCP.
Your raw recording and transcription stay on your device, so you can always go back to the original.
The prompt behind this context
I'm going to talk through my feedback on a code review / pull request. Turn it into organized review comments with four bold sections: **Blockers (must fix before merge)**, **Suggestions**, **Nits**, and **Praise**. Put each point under the right severity, keeping my technical specifics, file/line references, and function names exactly. End with an italic overall line if I give a verdict. Don't invent issues or upgrade a nit to a blocker. Omit any empty section. Output only the review.
Make it your own. This is a starting point. Once it's in Contextli, tweak the instructions so the output comes out exactly how you like it.
Use this context
One tap adds it to your clipboard. Open Contextli and paste to add it.
Next, open Contextli, Contexts, Import, paste.
Make it your own. This is a starting point. Once it's in Contextli, tweak the instructions so the output comes out exactly how you like it.
Your raw recording and transcription stay on your device, so you can always go back to the original.
Related contexts
Code Decision Log
Right after you make the call, while the tradeoffs are still fresh, talk through why you went this way and what you rejected. Three months from now when someone asks 'why a queue here?', the answer is already written down.
Standup Update
Standup's in two minutes and you're still reconstructing yesterday. Skip the typing. Talk through your day in whatever order it comes out, and get a clean yesterday, today, and blockers post you can drop straight in the thread.
Technical Spec Note
You've figured out the approach in your head and writing the spec feels like a tax. Talk it through, the problem, the approach, the tradeoffs, the open questions. You get a structured skeleton you can flesh out, instead of staring at a blank design doc.