I have gone through my pickaxe workflow, answered all the questions that are built into it and when it gets to the end, at which point it should be writing the article based on how I’ve trained it and based on the fact that it’s using the Claude 4.6 Sonnet model. It just says thinking, stops and doesn’t post anything else.
What’s going wrong please?
Hi @bencaine ,
Thank you for reaching out.
Could you please email us at info@pickaxe.co with a few more details so we can investigate this properly?
Please include the link to the Agent, the current token length limits you have configured, and whether this happens with other models as well or only with Claude 4.6.
In the meantime, I highly recommend increasing the Token Length settings for your Agent. You can find this by going to:
Configure tab → Token Lengths → Input and Output
If the input or output token length is set too low, the Agent may not have enough room to process the full workflow conversation and generate the final article. Since your workflow asks multiple questions before writing the article, the Agent needs enough input tokens to “remember” the previous answers and enough output tokens to write the completed response. If those limits are too low, the response can stop before finishing or fail to produce the final article properly.
