Copilot Claude Opus 4.8 emits pseudo tool-call text instead of structured tool calls

[doomsday616]

Description

With github-copilot/claude-opus-4.8, OpenCode can persist pseudo tool-call text as normal assistant text instead of receiving a structured tool call.

The malformed assistant message can look structurally like this:

• role = assistant
• providerID = github-copilot
• modelID = claude-opus-4.8
• finish = tool-calls
• structured tool parts: 0
• assistant text contains pseudo tool-call markup such as <invoke ...> or short call <tool> lines

This appears to be the root cause behind the assistant-prefill 400 tracked in #31807. After such a malformed finish=tool-calls turn, OpenCode continues the tool loop, but the conversation history effectively ends with an assistant message. Copilot Claude Opus 4.8 then rejects the follow-up request.

Steps to Reproduce

I do not yet have a small deterministic repro, but the issue occurred in real OpenCode TUI sessions with a tool-heavy workflow:

1. Start an OpenCode TUI session with github-copilot/claude-opus-4.8.
2. Perform a long multi-step task with many tool calls (read, write, edit, grep, bash, etc.).
3. Continue the same session through many tool-call turns.
4. Inspect local SQLite message / part metadata.
5. Observe assistant text parts containing pseudo tool-call text such as <invoke ...> or call read, while the assistant message has finish=tool-calls but no structured tool part.

Expected Behavior

• If the model requests a tool, OpenCode should receive or recover a structured tool call.
• Pseudo tool-call text should not be persisted as normal assistant prose.
• A response with finish=tool-calls should not leave OpenCode with zero actual tool parts and then continue the tool loop blindly.

Actual Behavior

In the affected session, Opus 4.8 produced malformed assistant turns where pseudo tool-call text was persisted as normal assistant text parts.

Same-session structural comparison:

Model	Assistant messages	<invoke> text parts	call ... text parts	finish=tool-calls with 0 tool parts	assistant-prefill API errors
claude-opus-4.6	161	0	0	0	0
claude-opus-4.7	71	0	0	0	0
claude-opus-4.8	337	13	229	6	6

Representative malformed turn shape:

model: github-copilot/claude-opus-4.8
role: assistant
finish: tool-calls
structured tool parts: 0
assistant text contains: <invoke ...>

Every sampled assistant-prefill 400 in #31807 was immediately preceded by this malformed shape.

Additional Evidence

Sanitized DB-derived metadata only:

https://gist.github.com/doomsday616/d3fb56acd0f14964a9797df35f1fcd96

The gist contains only structural metadata from local SQLite message / part rows. Private message text, local paths, usernames, WeChat identifiers, and conversation content are not included.

OpenCode Version

1.17.7

Operating System

Windows 11

Terminal

Windows Terminal / PowerShell 7

Provider / Model

github-copilot / claude-opus-4.8

I tested the same affected session metadata against other Copilot Claude models. In that session, claude-opus-4.6 and claude-opus-4.7 did not show the same malformed pseudo tool-call pattern.

Related Issues

• Copilot Claude Opus 4.8 assistant prefill 400 after malformed pseudo tool-call text #31807 — downstream assistant-prefill 400 after the malformed pseudo tool-call turn leaves the next request ending with assistant.
• gpt-5.3-codex intermittently leaks malformed tool-call text and stops mid-run #15164 — similar class: malformed tool-call text leaks into visible output.
• Interrupted streamed apply_patch call leaves orphaned tool state and leaks internal preparation text #30653 — related interrupted/orphaned tool state and leaked internal preparation text.
• Truncated tool calls are misclassified and unrecoverable (finishReason: length + repairToolCall + doom loop) #18108 — related malformed/truncated tool-call recovery area.

Sanitization

The original full session contains private content. The linked gist only includes sanitized structural metadata needed for debugging.

[github-actions]

This issue does not fully meet our contributing guidelines.

What needs to be fixed:

• The issue body is explicitly a placeholder: it states that real content "will be added in follow-up comments." Issue bodies must contain the actual reproduction details, not a promise to add them later.
• A bug report template was not used — required fields (Description, Steps to Reproduce, OpenCode version, OS, Terminal) are either absent or left empty.
• No steps to reproduce are provided, making it difficult to investigate the problem.

Please edit this issue to address the above within 2 hours, or it will be automatically closed.

---

This issue might be a duplicate of existing issues. Please check:

• This model does not support assistant message prefill / Github Copilot with Opus 4.6 #13768: Same error ("This model does not support assistant message prefill") with GitHub Copilot + Claude Opus — 70 comments of prior investigation
• Bedrock Claude Opus 4.6 rejects assistant message prefill #29760: Same prefill rejection error with Claude Opus 4.6 (different provider)

If you believe this was flagged incorrectly, please let a maintainer know.

[github-actions]

Thank you for updating the issue with the full report! It now meets our contributing guidelines. The needs:compliance label has been removed.

[rekram1-node]

can u sed the session this happened in? opencode export > session.json

[rekram1-node]

I need at least some info from the session just manually sanitize it if ur super concerned