Tracecat
Tracecat's 1.3 alphas keep reworking agent plumbing: tunnels, multi-org MCP, pluggable runners.
A side-by-side editorial comparison of Atlassian and Productboard — release velocity, themes, recent moves, and the top alternatives to consider.
Atlassian is productizing its own agent plumbing while pushing Rovo deeper into Microsoft's surfaces.
Atlassian's feed runs two parallel threads. One is distribution: Rovo and the Teamwork Graph are being pushed into Microsoft 365 and Teams, so Atlassian context and Jira actions reach users who never open an Atlassian tab. The other is internal engineering practice written up in public — agent harnesses, sandboxed execution, and pipelines that take routine work like vulnerability remediation off a human's queue. The stream stays mostly commentary; genuine releases are a minority and sit among culture posts and customer stories.
Productboard v2 adds exact-match text filters as the data model keeps widening.
Productboard's v2 API spent the summer closing v1 parity gaps, then in September made entity types an open, workspace-specific string. Since then the data model has kept widening: rich text accepts tables and checklists, relationships hold their own field values, and as of October 15 entity search supports exact-match filtering on text fields.
Atlassian's feed runs two parallel threads. One is distribution: Rovo and the Teamwork Graph are being pushed into Microsoft 365 and Teams, so Atlassian context and Jira actions reach users who never open an Atlassian tab. The other is internal engineering practice written up in public — agent harnesses, sandboxed execution, and pipelines that take routine work like vulnerability remediation off a human's queue. The stream stays mostly commentary; genuine releases are a minority and sit among culture posts and customer stories.
The two threads are converging on the same bet: that the durable asset is the graph and the agent runtime underneath, not any single surface. Recent posts move from describing agents that answer questions to agents that hold a task end to end — pick up a ticket, change code, open a pull request, close the loop after deploy. Publishing the architecture behind Rovo's agent harness alongside the internal case studies reads as groundwork for exposing that harness to customers rather than keeping it as an internal tool.
Expect the Agentic Pipelines work described in the Bitbucket posts to surface as a customer-facing Bitbucket capability, and the Teamwork Graph connector list to keep growing to more non-Atlassian sources. The entries here don't indicate timing or packaging.
Productboard's v2 API spent the summer closing v1 parity gaps, then in September made entity types an open, workspace-specific string. Since then the data model has kept widening: rich text accepts tables and checklists, relationships hold their own field values, and as of October 15 entity search supports exact-match filtering on text fields.
With parity mostly done, v2 is extending the data model and making it queryable. Workspace-defined entity types and relationship fields push schema decisions into each workspace's configuration, and filter operators like eq let integrations match records precisely instead of scanning with contains.
Expect filtering and search to reach relationship fields next, following the pattern of new data shapes gaining query operators a few weeks later. Whether the beta customer score endpoint leaves beta isn't signalled by the entries.
Other PM products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either Atlassian or Productboard.
Tracecat's 1.3 alphas keep reworking agent plumbing: tunnels, multi-org MCP, pluggable runners.
NocoBase's Runner Agent gets conversation management, task links and agent switching mid-thread.
Camunda opens the 8.11 cycle with job-worker fixes and variable metadata across gRPC, REST and MCP.
Todoist's 2026 bets are on capture: voice via Ramble, photos and files via Task Capture.
Atarim's AI experts now propose changes to live pages as drafts, and the app runs on phones.
Fibery's MCP server can now upload local files, so AI assistants can attach screenshots and images.
See all Atlassian alternatives → · See all Productboard alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Atlassian is currently shipping more aggressively (velocity 10.0 vs 6.3), with 0 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. Atlassian is currently shipping more aggressively (velocity 10.0 vs 6.3), with 0 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other PM products to evaluate alongside.
Top Atlassian alternatives in PM are ranked by recent ship velocity. Browse the "Atlassian alternatives" section above for the current picks, or visit /alternatives/atlassian for the full list with editorial commentary on each.
Top Productboard alternatives in PM are ranked by recent ship velocity. Browse the "Productboard alternatives" section above for the current picks, or visit /alternatives/productboard for the full list with editorial commentary on each.