Raw notes in,clean Linear issues out.
Bug reports arrive as screenshots, Slack shorthand, and half-sentences. Khint drafts the issue with AI — title, repro, acceptance criteria — and files it in Linear from one keyboard shortcut.
A vague issue costs more
than the bug itself.
“Login broken for some users, see thread” gets triaged last, bounced back with questions, and re-opened twice. Writing it properly — what happened, on which build, how to reproduce it, what done looks like — takes ten minutes you don't have when six reports land at once. So either quality slips, or your morning disappears into transcription. Khint attacks exactly that gap: its Agents draft the issue from whatever raw material you have, Capture turns screenshots into text first, and Memory keeps the triage-hour context attached to every draft.
One triage hour, start to filed issues.
A Tuesday-morning bug triage, step by step.
Connect Linear once
Paste your Linear API key into the Linear card on Khint's Integrations page. Khint fetches your teams live, and from then on any workflow can carry a Create Linear issue step pinned to the team you choose in the editor. How integrations work →
Start a session for the triage hour
Before you open the support thread, start a Memory session — “Tuesday triage”. Every capture and every Action you run gets logged, and Khint keeps a compact running summary. Each issue you draft later already knows what the previous ones were about. How Memory works →
Capture the evidence
The bug arrives as a screenshot of a stack trace and a blurry photo of an error dialog. Hit Cmd+Shift+K, pick Extract text, drag a region — the text is OCR'd to your clipboard. If you need to interrogate the screenshot first, pick Capture & askand ask questions about what's on screen. How Capture works →
Run the drafting workflow
Select the raw material — pasted OCR text, the customer's message, your shorthand — and launch the workflow from the palette. Step one tightens it into a reproducible bug report; step two adds acceptance criteria; step three is the Create Linear issue step. The first line becomes the title, the rest becomes the body. How Agents work →
Get the link back, keep the trail
The created issue's link is the workflow's output — paste it straight into the support thread. The filing is also logged into your active session, so at the end of the hour your Memory session reads like a triage report: what came in, what you filed, where it went.
The workflow is yours to shape.
Up to five steps in series — AI Actions plus an integration write at the end.
Swap the drafting Actions for your own — a stricter repro template, your team's definition of done, a translation step for a bilingual team. Or skip the hand-assembly: describe the flow in plain language and Khint's workflow generator builds the chain for you, reusing Actions you already have and slotting in the Linear step when the integration is connected. The same workflow machinery also writes to Jira and Confluence if your team lives in Atlassian instead.
Why not draft in a chatbot tab?
You can — and every issue becomes four round-trips: retype what the screenshot says, re-explain the project, copy the draft out, switch to Linear and paste it in. Khint collapses them into one run on the text you selected, with the screenshot already read, the context already attached, and the issue already filed. And if you want a longer working session with your AI client, Khint ships an MCP server so Claude Desktop or Codex can read your triage session directly — the captures, the drafts, the filed issues — without copy-paste.
Common questions.
Can AI create Linear issues for me?
Yes, as an explicit step you trigger — never in the background. Connect Linear with an API key from Khint's Integrations page, then add a Create Linear issue step to a workflow and pin the target team in the editor. When you run the workflow on selected text, the drafting Actions run first, the issue is filed in Linear, and the created issue's link comes back as the workflow's text — ready to paste into a thread.
How does Khint know the context behind the issue?
Through Memory sessions. Start a session for your triage hour and everything you capture or run is logged and compacted into a short summary. That summary rides along with every AI Action, so the drafting prompt already knows which release you're stabilizing and what was decided earlier — without you re-explaining it. Sessions live in a local SQLite database on your machine.
Do I have to write the drafting prompts myself?
No. The Action editor has a Generate button that drafts a prompt from a plain-language description, and an Improve button that refines an existing one. You can also describe the whole flow — "tighten the bug report, add acceptance criteria, file it in Linear" — and Khint's workflow generator assembles the chain, reusing your existing Actions where names match and proposing the Linear step when the integration is connected.
What does Khint send to the AI when I draft an issue?
Only the text you selected, plus — if a Memory session is active — the compacted session summary. Nothing else leaves your device silently. The Linear API key is stored in your OS keychain, and the issues Khint creates are logged into your local session history so you can trace what was filed and when.
Try it on tomorrow's triage.
Free with 10 AI actions and 5 captures per day. No credit card. Capture the screenshot, draft the issue, get the link back.