Repository navigation
[FEATURE]: A /injected-messages command that shows the full message array being sent to the model, including the system prompt, tool definitions, agent-change injected messages, and internal orchestration context #33333
Description
Feature hasn't been suggested before.
- I have verified this feature I'm about to request hasn't been suggested before.
Describe the enhancement you want to request
This could help debug problems that happen when changing agent or model during a conversation, for example. It could also help users understand and better learn how LLMs and agentic harnesses work.
I know that in day to day usage no one wants to see this stuff, but I think that a command used to enable it could be useful in some situations.
Metadata
Metadata
Assignees
Labels
No labels
Activity
github-actions commentedon Jun 22, 2026
This issue might be a duplicate of existing issues. Please check:
Rexarrior commentedon Jul 31, 2026
I opened a much smaller PR that covers only the system-prompt part of this request: #39905
It adds:
This does not dump the full message array, tool definitions, provider options, or all injected orchestration context. It only prints the resolved effective system prompt for the selected project/agent/model.
I think it can still be useful as a low-risk first debugging primitive for checking
AGENTS.md,opencode.jsoninstructions, skills, MCP instructions, custom agent prompts, andexperimental.chat.system.transformplugin changes.github-actions commentedon Sep 30, 2026
To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.