Author-written functions in Flow Designer

Verified 2026-09-05 · 13 claims · sources tier 2 · 1 disputed

Also known as Function activity, Function as a Service, Pebble Template.

Flow Designer functions run author-written code in supported runtime environments 9. They allow supervisors and administrators to execute custom logic directly within flow execution 9​12.

Runtimes and execution

Flow Designer functions run author-written JavaScript on Node.js 22.x and Python 3.13, which are described as available in the first phase 9. Function code is written with language-specific static analysis, but the documentation never states what that analysis checks for 11.

The only sentence characterising the function execution environment names a packaging and invocation model rather than a network policy 3. Furthermore, no sentence in the functions documentation or its frequently asked questions uses the words network, internet, outbound, egress, sandbox, isolation, firewall, allow-list, or proxy 5. Regarding supported dependencies, the complete published statement on what code inside a function may use is that common native libraries are available and custom modules are not supported; the native libraries are never enumerated 4.

Configuration and lifecycle

Supervisors as well as administrators can manage functions, gated by a Function setting in the user profile that can be set to Edit, View, or None 12.

Publishing a function after editing its source code or settings updates the change immediately in every flow consuming those tagged versions, and the only stated safeguard is an instruction to be mindful 7. When debugging code, console logs are not exposed in the function test runner; a function can surface debug information only by setting and reading a custom output string array 2.

Published limits

Functions carry 6 published limits: 200 functions per organisation, 128 MB of memory per function, 5000 lines of code, 5 seconds of execution, 10 input variables, and 1 JSON output variable 8. However, sources disagree on the output variable capacity: one frequently asked question states that a function sets 1 JSON output variable and another states that multiple variables can be set from the code block 6.

The system limits article contains no occurrence of the word function anywhere in its text, tables, or headings, so it corroborates none of the published function limits 13. In addition, neither the Flow Designer article's self-loop table nor the system limits article's copy lists Function among the throttled activities, so no published ceiling caps how many times a function can be re-entered in one interaction 1. Across all 21 activities that are listed, the self-loop limit tables in the Flow Designer article and the system limits article agree on every value 10.

See also

• The Webex Contact Center Flow Designer activity set
• Webex Contact Center system limits reference

Applicability

Across the claims in this article, the evidence covers Cisco Webex Contact Center, verified as of 2026-09-05. The deployment model is not covered by any claim in this article.

What remains uncertain

Whether function code can open a network connection is not covered by any claim in this article.

Which native libraries are available inside a function is not covered by any claim in this article.

What the language-specific static analysis checks is not covered by any claim in this article.

Whether an exported function JSON carries anything beyond source is not covered by any claim in this article.

WarmTransfer. (2026, September 5). Author-written functions in Flow Designer. WarmTransfer. https://warmtransfer.net/knowledge/wxcc-flow-functions

@misc{warmtransfer-wxcc-flow-functions,
  title  = {Author-written functions in Flow Designer},
  author = {{WarmTransfer}},
  year   = {2026},
  url    = {https://warmtransfer.net/knowledge/wxcc-flow-functions},
  note   = {Verified 2026-09-05}
}