[bug] web retrieval step fails with validationerror but downstream seo step still succeeds

[rickpixelml]

Summary

When running the "SEO Meta Writer" workflow from a workspace API-key-triggered history run, the "Fetch page content" step fails with a Pydantic validation error because the web retrieval node returns None for WebRetrievalNodeOutput.result instead of a string. The workflow then continues to "Generate SEO meta", which produces output without the fetched page content.

Observed behavior

• Input URL: https://posthog.com/blog/what-is-product-analytics
• Primary keyword: product analytics
• Brand name: PostHog
• "Fetch page content" fails after ~3.20s with:

Step Failed ValidationError
1 validation error for WebRetrievalNodeOutput
result
  Input should be a valid string [type=string_type, input_value=None, input_type=NoneType]

• "Generate SEO meta" still reports success and says it generated output based on the URL and primary keyword because the page content was not provided.

Expected behavior

If the web retrieval result is missing/None, the node should either:

• normalize the result to a valid empty string and expose a clear no-content state, or
• fail the workflow in a way that prevents downstream nodes from treating missing content as a normal successful input.

Notes

The run URL contains workspace/workflow/log identifiers, so I am not including the full URL publicly. Screenshot attached below.