Jira joins the connectors
The Jira connector lands — projects and issues as retrievable knowledge, hourly incremental — alongside per-source reconcile schedules: a second full-sync cadence, on every connector, that prunes what incremental syncs can never see deleted.
A ticket tracker is the system of record for what work exists and what state it's in — which makes it
exactly the kind of source an agent team should be able to read. But Jira is also a system where
things get deleted, and an incremental sync can never notice: a deleted issue simply stops matching
updated >= the watermark. v2.2.0 ships both halves — the connector, and the mechanism that keeps
any connector's index honest.
What's new
- Jira connector. Ingests projects and issues (key, summary, status, description as plain text)
from a Jira Cloud site via
POST /rest/api/3/search/jql, the current search endpoint. Auth is an Atlassian email + API token, encrypted in the vault. Projects are opt-in viaprojectKeys— only what you list syncs. Incremental runs use a relative-minutesupdatedwindow (timezone-proof, with a 5-minute overlap that content-hash dedup makes free), and issues are keyed by the immutable numeric id, so moving one between projects doesn't duplicate it. Done detection ridesstatusCategory, with anotDoneStatusesoverride for admins who park "Won't Do" in the done column. 429s honorRetry-After; an expired token surfaces as an actionable "reconnect" error, not a retry loop. - Per-source reconcile schedules — on every connector. The bigger half. A source's
schedulecron drives incremental syncs; a new second cadence,reconcileSchedule, drives a periodic full sync whose tombstone pass prunes records deleted upstream. Omit it and the connector's default applies (Jira reconciles nightly at 03 UTC); setfalseto disable. Both are durable Temporal Schedules on the same ingestion pipeline. - Config changes take effect now.
workspace:applystarts a one-off full sync when a source's config changed — widenprojectKeysand the new project starts pulling immediately; narrow it and the out-of-scope documents get pruned, instead of waiting for the next reconcile.
Configuration
slug: jira
name: Jira — Projects & Tickets
kind: jira
config:
baseUrl: https://acme.atlassian.net
projectKeys: [REV, OPS] # opt-in — only listed projects sync
doneWindowDays: 90 # full syncs keep done issues updated in this window
schedule: '17 * * * *' # hourly incremental
# reconcileSchedule: '0 3 * * *' # full-sync reconcile — omit for the connector default, false to disable
enabled: true
Apply the workspace, paste the Atlassian email + API token on /dashboard/sources, and add jira
to connectorSources on the agents that should read it.
Where to find it
- Release: https://github.com/vocion/vocion-core/releases/tag/v2.2.0
- Code:
packages/core/src/libs/sources/jira.ts,packages/core/src/services/SourceScheduleService.ts - Docs: Connectors & data sources, Connecting sources