// verdict · checked 2026-08-11 · Gumloop agents (docs.gumloop.com, 2026)
Gumloop
Gumloop
Drag, drop, call it autonomous.
- CAPABILITY
- 4/5 AGENT
- PROOF
- OPERATOR INSPECTABLEYou can run it and read the steps yourself — the record exists for whoever runs it
- TRACE ACCESS
- OPERATOR-OWNEDWhoever runs it holds the record; nothing is published
- LEASH
- Not measuredno execution record scored — measure it yourself
| LOOPSruns until doneyes | Agent "decides which tools to use and in what order, runs them, adapts based on the results"; Max Steps caps tool calls before it must respond (default 100, range 1-200) — a ceiling, not a fixed script. |
| CHOOSESpicks its own toolsyes | Docs contrast flows' fixed path with agents that at runtime analyze the request and decide which tools to use and in what order, including when to invoke a flow as a tool. |
| ACTSchanges the worldyes | Code Sandbox is "natively enabled on all agents with no configuration required" and runs Python and shell commands, creates files, and calls connected integrations via the gumloop SDK. |
| RECOVERSfixes its own errorsno | Only model-outage retry/fallback is documented; docs tell the user to review tool errors themselves. |
| UNSUPERVISEDnobody watchingyes | Tool Management preset "Always allow" executes tools without pausing, and schedule/app-event triggers start the agent unattended; note docs never state Always allow is the default preset. |
Capability decisions combine documented behaviour and execution evidence. The proof label on each row shows which of the two decided it, and Leash is reported separately, only when it was measured from an execution record. Primary source: docs.gumloop.com. Tested against: Gumloop agents (docs.gumloop.com, 2026). Our confidence in this dossier: 8/10.
Can you see it work?
the weakest of the five grades aboveOPERATOR INSPECTABLE — You can run it and read the steps yourself — the record exists for whoever runs it. Trace access: OPERATOR-OWNED — Whoever runs it holds the record; nothing is published.
Run Log per execution, run-history API, checkpoints, enterprise audit logs
Where we looked: docs.gumloop.com/llms.txt index: Run Log core concept, checkpoint history, retrieve-run-history and retrieve-run-details API endpoints, enterprise audit logging. · see the evidence →
- 2026-08-113/5 WRAPPER WITH AMBITION → 4/5 AGENTLOOPS corrected to yes: Docs: agent decides which tools and in what order, loops until done or blocked; Max Steps (default 100, max 200) is a ceiling, not a fixed count.editorial correction — no case was filed
Badges
earned, never for sale
These render from the live dataset, so a badge cannot outlive the verdict it claims. If the verdict changes the badge changes with it, and an unearned one returns 409 rather than an image. No sponsor can buy one, at any price.
Think this is wrong?
bring evidence, not opinionName the criterion and link a trace, a doc, or a recorded run that shows the behaviour. A verdict changed by evidence is the best thing that can happen to this site, and the change goes on the public record with your reason attached. Vendors are welcome, and a vendor's own filing is labelled as such rather than buried.