fix: strip tool history from Write content-generation context#777
Merged
bobleer merged 1 commit intoMay 19, 2026
Merged
Conversation
Weak models imitate tool-call XML format when the history passed to the second-stage content-generation request contains tool_use/tool_result messages, causing raw XML to be written into the file instead of actual content. - Drop role=tool messages from content_messages entirely - Replace assistant messages that carry tool_calls with a plain-text summary [called tools: ToolName] so the model sees no tool-call syntax - Add prompt rule 7 explicitly forbidding tool_call XML inside <bitfun_contents> tags
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
Weak models imitate tool-call XML format when the history passed to the
second-stage Write content-generation request contains tool_use/tool_result
messages. This causes raw XML to be written into the file instead of actual
file content, triggering a Delete -> Write cycle from the user's perspective.
Root cause
In
generate_write_tool_contents,content_messageswas built by directlycloning the full
ai_messageshistory, including all tool-call and tool-resultturns. Weak models see this format in context and reproduce it inside
<bitfun_contents>tags.Fix
role=toolmessages fromcontent_messagesentirelytool_callswith a plain-text summary[called tools: ToolName]so no tool-call syntax appears in context<bitfun_contents>tagsVerification
cargo check -p bitfun-corepasses.