AE-FBZ9M6NV
Stateful agent shows inner monologue but no user reply because the model response lacks the send_message tool call; fix did not resolve it for users
TOOL FAILUREseverity: HIGHcause: UNKNOWNoutcome: UNRESOLVEDconfidence: LOW
Letta agents on OpenAI models frequently hit "API call didn't return a message": the response contained no send_message tool call, so only inner monologue was shown. A maintainer patch to the inner-thoughts unpacker was merged, but two users reported the problem persisted on the fix and nightly builds, and one later hit a 400 about a null message content on 0.6.7.
- Framework / agent
- Letta · stateful memory agent (MemGPT-style) using a send_message tool
- Remediation attempts
- FAILED
- Recurrence
- not documented
- Source languages
- en
- Updated
- 2026-09-29
Sources
- GITHUB ISSUE "API call didn't return a message" error repeating very frequently in v5.5, with openai. — github.com/letta-ai/letta, retrieved 2026-09-29
- GITHUB PULL REQUEST fix: patch bug in inner thoughts unpacker — github.com/letta-ai/letta, retrieved 2026-09-29
Symptoms
- The UI shows the inner monologue but the agent never sends a user-facing reply
- Response choices contain no function call when the error occurs
- Existing agents are affected on 0.6.5 while new agents work
- On 0.6.7 every step fails with a 400 because a message content is null
The full record — root-cause evidence, every remediation attempt with its status and verification, failed attempts, patch references, verbatim quotes and recurrence — is a paid lookup (0.018 USDC via x402). Agents:
GET /api/v1/cases/AE-FBZ9M6NV. PricingSimilarity to your system is not implied. A remediation that worked in the documented context may not work in yours.