Who this is for
Developers and on-call engineers who need an accurate incident record before adrenaline erases the timeline.
The moment this saves you
I resolve a production incident at 3am running on adrenaline, and by the postmortem I can't reconstruct the exact order of what I tried or which fix actually worked, so the writeup is guesswork.
See it work
Messy spoken thought in. A clean, structured artifact out.
Okay incident just resolved, capturing while it's fresh. Around 1:40am the API started throwing 500s, error rate spiked to like 40%. First thing I did was check the dashboards, saw the database connection pool was maxed out. I tried restarting the API pods, that helped for like two minutes then it came back, so that wasn't it. Then I realized a migration that deployed at 1:30 had added a query without an index, that was hammering the database. I rolled back the migration and the error rate dropped within a minute. So root cause was the missing index. Total downtime about 25 minutes. Follow up is we need index review in the migration checklist.
Incident, June 5, 2026
- Impact: API returning 500s, error rate spiked to ~40%, ~25 min of degradation
Timeline
- ~1:40am, alerts fired; dashboards showed the DB connection pool maxed out
- Restarted API pods, recovered ~2 min then regressed (not the fix)
- Identified a 1:30am migration that added an unindexed query hammering the DB
- Rolled back the migration, error rate dropped within a minute
- Root cause: Missing index on a query added by the 1:30am migration
- Follow-up: Add index review to the migration checklist
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 a production incident I just handled, probably out of order and adrenaline-fueled. Turn it into a postmortem-ready note: a bold "Incident, [today's date]" heading, an Impact line (what broke, severity, rough downtime), a bold **Timeline** as a NUMBERED list reconstructing what happened and what I tried in order (mark steps that did NOT work), a Root cause line, and a Follow-up line for prevention. Keep my exact times, error rates, and technical details. Mark anything I'm unsure about as such; don't invent steps, times, or the root cause if I didn't land on one. Output only the note.
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 click copies it and shows you exactly how to drop it into Contextli.
Next, open Contextli, go to the Contexts page, click Import, choose From JSON, paste, then Import Context. It is ready to use.
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.
Sprint Retro Notes
The retro generated real insight and then it died in a doc nobody reopens. Right after, talk through what went well, what didn't, and what you'll change. You get a structured retro with owned action items, so next sprint actually improves on this one.
Product Bug Report
You just watched something break mid-demo. Don't lose the repro hunting for the tracker. Mutter what happened, what you clicked, what you expected, and walk away with a report engineering can actually act on.
Questions people ask
Questions developers ask about On-Call Incident Note
What should an incident report include?
A good incident report includes a clear title, steps to reproduce the issue, the expected behavior, the actual behavior observed, the severity or priority, and the environment details such as OS, browser, and app version. The On-Call Incident Note context structures your spoken description into these fields automatically, so nothing gets left out when you are in the middle of debugging.
How do I write an incident report in under a minute?
Speak what you found: describe the issue, what you expected, what actually happened, and how bad it is. The On-Call Incident Note context structures your words into a complete incident report you can paste directly into Jira, Linear, or GitHub Issues. Most take under 60 seconds to dictate, so you capture them without breaking your flow.
How do developers capture issues without interrupting their flow?
The key is to capture the issue immediately without switching context mentally. Contextli lets you speak a quick voice note describing it and produces an incident report from it. You can dictate while the issue is still on screen, then paste the formatted output into Jira, Linear, or GitHub Issues when you come up for air. No typing is required during the capture step.
Can I write an incident report by talking instead of typing?
Yes. The On-Call Incident Note context lets you speak a description in plain language and converts it into a structured incident report with all the required fields filled. You speak the way you would explain it to a colleague, and the context handles the formatting.
How do I add this context to Contextli?
Copy the context on this page, then open Contextli and go to the Contexts page. Click Import, choose From JSON, paste it into the Import from Clipboard window, and click Import Context. It is ready to use in under 30 seconds. If you do not have Contextli yet, you can download it for free first.
Is my voice recording private? Does Contextli send it anywhere?
Your voice recording and the transcription are stored on your device only. Contextli processes your audio locally and does not send your recordings or transcription text to any server. The structured output it produces is text you control, and you decide where it goes.
Can I change what the output looks like?
Yes. Every context in Contextli is a starting point you can edit. Open the context in the app, change the instructions to adjust the structure, tone, or fields, and save. The next time you use it, the output reflects your changes. You are not locked into the default format.
Do I need to install an app to use this context?
Yes. Contextli is a free app. Download it, then copy this context and paste it into the Import from Clipboard window on the Contexts page. The whole process takes about 30 seconds.