Solved it once? Make ita knowledge base article
Every resolved ticket that never becomes an article is the same question answered again next month. Khint pulls the resolved Jira issue in, drafts a customer-facing article with the template you saved, and publishes it to Confluence, from one keyboard shortcut, without retyping what the resolution notes already say.
The write-up is always
the step that gets skipped
Nobody skips it out of laziness. Turning a resolved ticket into an article means re-reading the thread, translating engineer shorthand into customer language, forcing it into the KB template, and pasting it into Confluence, a 25-minute chore per ticket that always loses to the live queue. So the same question comes back, and the answer lives in one senior agent's head. Khint compresses the chore with its three pillars, Agents, Capture, and Memory, wired together behind one palette: the ticket comes to you, the template is a saved prompt, and publishing is one more row in the same list.
From resolved ticket to published page
A Friday-afternoon KB pass, one ticket at a time.
Start a session for the KB pass
You have blocked Friday afternoon to pay down the write-it-up-later debt. Start a Memory session, call it “KB pass”. Everything you pull and draft this afternoon is logged to it, and its compact summary rides along on each AI call, so by the third article the model already knows which product area and release you are documenting.
Pull the resolved ticket in
Hit Cmd+Shift+K and run the Jira Pull issue row: a picker lists your recent issues, so the resolved ticket is one click away. Its summary, status, and description land in your session. If the actual fix was hashed out elsewhere, the Pull thread rows for Slack and Gmail bring that conversation in too. How integrations work →
Save your article template as an Action
Write the KB house style down once as a saved Action: “Draft a KB article with Symptom, Cause, Fix, Applies to. Address the reader as the customer. No internal jargon, no customer names.” No prompt yet? Type that description and press Generate, then tighten it over time with Improve.
Draft the article where you work
Select the resolution notes, engineer shorthand and all, and run the Action. Back comes a customer-facing article in your structure, drawn from the selected text plus the ticket context already in your session. It pastes right back where you were, so you edit it in place: fix the version number, soften a sentence, done.
Publish it as a Confluence page
Run the Confluence Create page row on the finished draft and choose use the text as-is: the article becomes a page in your KB space, one page per run. Doing this weekly? Chain the drafting Action and the Create page step into a single Workflow, and next Friday each ticket takes one keystroke. How workflows work →
Resolution notes in, article out
The structure comes from the prompt you saved: this is one example of what a KB Action might return.
The template is not built into Khint; it is the Action you wrote, so a team that documents by product area, a team with a strict style guide, and a team writing for internal agents each get their own house structure from the same shortcut. The draft always pastes back to you first, and only the Confluence Create page run publishes it. This is a different job from drafting a reply to one customer or triaging the inbound queue: here the ticket is already closed, and the goal is that nobody has to ask again.
Why not paste the ticket into a chatbot?
You can, and for one article it works. For a weekly KB pass you will re-paste the template every chat, get a slightly different article shape each time, and copy every result into Confluence by hand. Khint pins the structure: the KB Action is written once and runs identically on every ticket, Memory carries the pulled issue and threads into each call so you are not re-explaining context, and the Jira and Confluence integrations mean the ticket comes in and the page goes out without a browser tab. And because an Action only sees the text you selected plus the session summary you opted into, resolved tickets full of customer context are not accumulating in a chat history.
Common questions
How does Khint turn a resolved ticket into a knowledge base article?
In three moves, all behind one shortcut (Cmd+Shift+K). Pull the resolved Jira issue into your Memory session so the resolution is in front of you. Run a saved AI Action, your house template for a KB article: symptom, cause, fix steps, applies-to. Then run Confluence Create page on the draft and the article lands in your space. The drafting prompt is yours, so every article comes out in the same structure regardless of which engineer wrote the resolution notes.
The real fix was discussed in Slack, not in the ticket. Can Khint see that?
Yes, if you pull it in. The palette has Slack Pull thread and Gmail Pull thread rows: run them on the conversation where the workaround was actually figured out and each lands as an entry in your active Memory session. Khint keeps a compact summary of that session and passes it along with your drafting Action, so the article reflects what actually fixed the problem, not just the two-line resolution field.
Does the article go straight to Confluence, or do I review it first?
You review it. The drafting Action pastes the article back where you are working, as text, and you edit it like anything else. When it reads right, run the Confluence Create page row on it; the create flow shows what will be sent and offers "Use the text as-is" so nothing is reworded on the way out. One page per run, nothing published silently.
Can I make this a one-keystroke workflow?
Yes. In the Workflow editor, chain your drafting Action with a Confluence Create page step, with the target space pinned in the step's settings. Select the resolution text, run the workflow from the palette, and the draft is written and filed as a page in one pass. Workflows take up to five steps, so you can add a second Action, say, one that strips customer names, before the page is created.
Where does the customer's data in the ticket go?
An Action only ever sees the text you selected, plus the compacted summary of your active Memory session if you opted in. Memory sessions live in a local SQLite database on your machine and are never synced to Khint's servers, and your Atlassian credentials sit in your OS keychain. If tickets carry emails or phone numbers, flip the per-Action redact PII switch and they become placeholders before the model sees them.
Close the ticket, keep the answer
Free with 10 AI actions and 5 captures per day. No credit card. Save the KB Action once, publish an article per resolved ticket after.