Sulu
Sulu ships 2.6 and 3.0 in lockstep, and the 3.0 line still reads like a maintenance branch.
A side-by-side editorial comparison of Gumloop and Pimcore — release velocity, themes, recent moves, and the top alternatives to consider.
| Feature | Gumloop | Pimcore |
|---|---|---|
| Sector | Mkt Auto | Mkt Auto |
| Velocity score | 7.5 | 5.0 |
| Sparks · 30d | 1 | 0 |
| Top themes | agent platform, company brain, mcp connectors, skills as code | pim-dam, security-hardening, sql-injection, backport-branches |
| Last editorial update | 2d ago | 2h ago |
| Website | — | Visit → |
Gumloop is turning its whole connector catalog into agent memory, one week at a time.
Gumloop launched Company Brain in early July and has widened it on almost every release since — Zendesk as a source, then permission-preserving Google Drive, then any connector in the catalog at all. The agent layer is picking up its own plumbing in parallel: organization skills sync from a GitHub repository and update as commits land, skills and connectors are manageable through the developer API (including by other agents), and agents can be DM'd directly in Slack. Models rotate through continuously, with Claude Opus 5 and Kimi K3 the latest, alongside bring-your-own Fireworks keys for running open models on a customer's own account.
Nearly every release in this window carries security hardening — Pimcore is working through an injection sweep.
Pimcore maintains two calendar-versioned lines in parallel, 2026.2.x and 2026.1.x, with fixes landing on the newer line and selectively backported. The content is dominated by input-validation work: explicit field allowlists on Custom Report updates, an ORDER BY column allowlist for redirect listings, rejection of unwhitelisted filter properties, escaped field names in DataObject JOIN conditions, hardened unserialize paths, and CSV exports bound to the requesting user. Feature work is thin by comparison — a CDN integration and a TemplateProviderInterface extension point in 2026.2.1.
Gumloop launched Company Brain in early July and has widened it on almost every release since — Zendesk as a source, then permission-preserving Google Drive, then any connector in the catalog at all. The agent layer is picking up its own plumbing in parallel: organization skills sync from a GitHub repository and update as commits land, skills and connectors are manageable through the developer API (including by other agents), and agents can be DM'd directly in Slack. Models rotate through continuously, with Claude Opus 5 and Kimi K3 the latest, alongside bring-your-own Fireworks keys for running open models on a customer's own account.
Two loops are closing at once. On knowledge, every connector Gumloop adds now doubles as a retrieval source, so catalog growth and Brain quality compound instead of competing for roadmap. On capability, skills are becoming versioned artifacts that live in a repository and agents are becoming things other agents can configure through the API — the platform is moving from a place you build one agent to a place agents are provisioned programmatically. Governance is following a step behind rather than leading: audit logs for Brain activity, access requests routed to approvers, per-conversation credit breakdowns, and usage-based billing for workflow AI steps.
Expect the same generalization that took Brain from fixed sources to any connector to be applied to triggers and actions, and expect the agent-configures-agent API to grow into templated provisioning so organizations can stamp out fleets rather than assemble agents one at a time.
Pimcore maintains two calendar-versioned lines in parallel, 2026.2.x and 2026.1.x, with fixes landing on the newer line and selectively backported. The content is dominated by input-validation work: explicit field allowlists on Custom Report updates, an ORDER BY column allowlist for redirect listings, rejection of unwhitelisted filter properties, escaped field names in DataObject JOIN conditions, hardened unserialize paths, and CSV exports bound to the requesting user. Feature work is thin by comparison — a CDN integration and a TemplateProviderInterface extension point in 2026.2.1.
The pattern is a systematic pass over places where user input reaches SQL or deserialization, spread across releases rather than bundled into one advisory — Custom Reports alone are touched in three separate releases. Alongside it runs a slower cleanup of the v11 era: admin functions removed, admin translations deprecated, v11-specific workflows refactored to v12. A task in the most recent release adds backward-compatibility break detection to the project's own code review tooling, which reads as an attempt to keep that cleanup from breaking integrators unannounced.
Expect the hardening sweep to keep producing small paired releases on both lines, with Custom Reports and listing filters the likely remaining targets, and further admin-layer removals as the v12 cleanup continues.
Other Mkt Auto 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 Gumloop or Pimcore.
Sulu ships 2.6 and 3.0 in lockstep, and the 3.0 line still reads like a maintenance branch.
The v4 canary line is being renamed Build Awesome while 3.x is kept on pure maintenance.
Steady feature-then-patch cadence, with security reports rising because AI tools are finding them
Measurement gets granular while the API quietly opens ClickFunnels up to outside agents
n8n's daily patch train hides the real work: hardening its MCP server into a proper auth resource.
OneSignal is arguing a new category into existence before showing the product behind it.
See all Gumloop alternatives → · See all Pimcore alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Gumloop is currently shipping more aggressively (velocity 7.5 vs 5.0), with 1 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. Gumloop is currently shipping more aggressively (velocity 7.5 vs 5.0), with 1 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Mkt Auto products to evaluate alongside.
Top Gumloop alternatives in Mkt Auto are ranked by recent ship velocity. Browse the "Gumloop alternatives" section above for the current picks, or visit /alternatives/gumloop for the full list with editorial commentary on each.
Top Pimcore alternatives in Mkt Auto are ranked by recent ship velocity. Browse the "Pimcore alternatives" section above for the current picks, or visit /alternatives/pimcore for the full list with editorial commentary on each.