Integration

Jira holds the project record. Cognia makes it answerable.

Issues, sprint plans, custom fields, comment threads. Cognia indexes them and connects them to the Slack threads and GitHub PRs that reference them.

What gets captured

●Issues (summary, description, comments, custom fields)

●Sprints and boards

●Project metadata

●ADF rich-text bodies (flattened to plain text for retrieval)

1
Open Admin → Integrations
Find Jira.
2
Atlassian OAuth (3LO)
We use Atlassian's OAuth 2.0 3LO flow. You'll pick a Jira site (cloudId is captured automatically) and approve the scopes.
3
Webhooks register on install
Cognia registers webhooks for issue events on the connected site. Updates land in retrieval within seconds.
4
Ask Jira-shaped questions
'What's open in the Polaris epic?' or 'Which P1 tickets does Sarah own?'

cogniahq.tech /agent · Jira

ask Cognia:What's the status of the Polaris epic?

Returns the epic, its open and closed children, the most recent comment thread, and the linked Slack discussion.

1Jira sourcejira

2linked memory meshgraph edge

cogniahq.tech /agent · Jira

ask Cognia:Why was COG-114 deprioritised?

Pulls the Jira ticket, the Slack thread where the deprioritisation was discussed, and the customer commitment that triggered it.

1Jira sourcejira

2linked memory meshgraph edge

read:jira-workread:jira-usermanage:jira-webhookoffline_access

Server vs Cloud?

Cloud only at the moment. Server / Data Center support is on the roadmap; ping us if you need it.

What's in the requested scope?

read:jira-work, read:jira-user, manage:jira-webhook, offline_access. No write scopes.

Multiple Jira sites?

Each site connects separately. You can have several Jira sites linked to one Cognia org.