Dimensions
Total Executions (per window)
- What: The count of Action/Flow runs initiated within a rolling window.
- Why: Protects the platform from unbounded load.
- Symptoms: New runs are rejected until the window resets.
Concurrent Executions (max in‑flight)
- What: The number of executions running at the same moment.
- Why: Prevents saturation and keeps latency predictable.
- Symptoms: Additional runs queue or fail fast with a concurrency error.
Max Execution Duration (per run)
- What: The upper bound on runtime of a single Action/Flow.
- Why: Avoids hangs and runaway costs.
- Symptoms: The run is terminated when the limit is reached.
Flow Depth (graph depth per run)
- What: The number of sequential Node steps allowed in one Flow execution.
- Why: Encourages composition and prevents overly deep graphs.
- Symptoms: Validation error before execution if the graph exceeds the limit.
Recursion Depth (self/nested re‑entry)
- What: The maximum allowed nested calls (Flows calling Flows/Agents recursively).
- Why: Stops infinite recursion and explosion of work.
- Symptoms: The triggering call is blocked beyond the threshold.
Practical strategies
- Batch & Bound: Break large inputs into lists; toggle Parallel on the consumer Node.
- Short‑circuit: Validate inputs early and exit fast on no‑op paths.
- Compose shallow: Prefer sub‑Flows for reuse but keep individual run paths short.
- Time‑box: Use conservative duration expectations in Actions; handle partial progress gracefully.
note: Reproduction aid — Save Payloads that reflect your highest‑throughput and longest‑duration cases so you can validate changes against quota‑sensitive scenarios.