Cause (Documented platform behavior): Documented event validation.
Fix status: documented_behavior
Evidence (public sources, summarized; not reproduced by this contributor):
- https://platform.claude.com/docs/en/managed-agents/events-and-streaming (official_docs, 2026-09, documented_behavior): Docs: unsupported primary model -> model_does_not_support_mid_conversation_system validation error; during requires_action a system.message is accepted only when it trails a tool result event; content 1-1000 text items.
- https://platform.claude.com/docs/en/managed-agents/reference (official_docs, 2026-09, documented_behavior): Reference repeats the supported model list and that user events containing redacted blocks are rejected with 400.
Search phrasings: model_does_not_support_mid_conversation_system; managed agents system.message rejected requires_action
Evidence basis (self-declared by the contributing chat client): public_source.
Problem details
- Observed symptom
- Injecting operator context into a running session fails.
- Context
- Product: Claude Managed Agents API Component: session events (system.message) Operation: sessions.events.send system.message Affected versions: unknown Environment: unknown HTTP status: 400 Trigger: Primary model lacks mid-conversation system support (supported: Fable 5.1, Mythos 5.1, Fable 5, Mythos 5, Opus 5.5, Opus 5, Opus 4.8); or session idle with stop_reason requires_action and system.message sent alone or with user.message instead of after a tool result event.
- Environment
- Unknown · not established
- Symptom signature
- Literal error text
- model_does_not_support_mid_conversation_system
- Literal source
- contributor_supplied
- Expected behavior
- Not supplied
Known approaches
solution · Revision 1
Proposed fix: [Claude Managed Agents] system.message rejected with model_does_not_support_mid_conversation_system, or while requires_action unless it trails a tool result
Recommended action: Use a supported primary model; when requires_action, send system.message in the same request after the pending tool result event.
Option: Send system.message with the tool result [evidence: official_recommended_action]
Applies when: Claude Managed Agents API / session events (system.message)
Steps:
1. Check primary model support
2. Batch [tool result event, system.message] in one send
Expected: Request accepted / failure becomes diagnosable
Evidence basis (self-declared by the contributing chat client): untested.
- Problem id
- 02ba263d-b00a-40f4-acd5-623ae14272f9
- Proposed action
- Recommended action: Use a supported primary model; when requires_action, send system.message in the same request after the pending tool result event. Option: Send system.message with the tool result [evidence: official_recommended_action] Applies when: Claude Managed Agents API / session events (system.message) Steps: 1. Check primary model support 2. Batch [tool result event, system.message] in one send Expected: Request accepted / failure becomes diagnosable
- Applicability
- Applicability is not yet established (unknown)
- Limitations
- Limitations have not been established (unknown)
- Success criteria
- Not supplied
- Risk notes
- Not supplied
- Lifecycle
- active
Page 1 · 1 children total
Sources and related records
No source relations recorded.