Pre-deploy reliability check for Kestra flows

Your flow authors beautifully. Will it survive its first failure?

Paste a flow's YAML and get a graded reliability scorecard in a second — with specific, one-line fixes. Every check maps to a real, documented Kestra failure mode, so you catch the fragility before production does.

✓ Runs entirely in your browser✓ Nothing uploaded, no account✓ Open methodology, issue-cited

Try it:

Runs in your browser · nothing is uploaded · ⌘↵ to check

The five dimensions

Failure handling

25 pts

Will a transient error take the whole run down? Checks retry coverage on fallible tasks, flow-level error handling, and a flow timeout.

#1493#1246

Observability readiness

20 pts

When this flow fails, will you be able to see why? Flags loop/subflow patterns whose failures surface poorly today, and fragile expression access.

#13115#16495#16373#11809

Scale risk

20 pts

Will this hold up under load? Flags high-frequency triggers paired with heavy work (worker-memory risk) and sequential-backfill exposure.

#13129#6129#2041

Guardrails

20 pts

Can a single execution run away or stampede? Checks per-task timeouts on long-running work and flow concurrency limits on triggered flows.

#1493

YAML hygiene

15 pts

Is the YAML itself a trap? Detects zero-width/invisible characters, tab indentation, and structural issues that silently break flows.

#13079

See the full rule list and the issue each one maps to in the methodology.

Built for where Kestra is going: agent orchestration

Kestra 2.0 orchestrates AI agents alongside data and infra. Agent tasks bring new failure modes — provider rate-limits, runaway calls that burn tokens, and cost stampedes when you loop an LLM over a list. Flow Doctor checks for those too. The thesis Kestra states out loud: the more autonomous a system becomes, the more engineering rigor it requires. Load the agentic flow example above to see it.