Repository navigation
ask: hide and retry a repeated system prompt (w3bs 0.1.20) - #278
Merged
Merged
Conversation
Live, the 0.5B strategist sometimes answers with its system prompt word for word, even with the question-first prompt (dev2 node, 2026-11-02). Now: - the CLI holds streamed tokens while they could still be that echo, so it never reaches the screen; - an answer that is only the echo is asked once more with the background in the system message (layout 1), on the CLI and on /api/v1/ask + w3bs_ask; - tests: fakes whose odd jobs echo, checking the retry and its layout. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Live verification of 0.1.19 hit the 0.5B strategist answering with its system prompt word for word (on dev2's node, with the question-first prompt). With this change:
/api/v1/askandw3bs_ask.I probed layouts on the same GGUF in Ollama: 5 questions at temperatures 0.2 and 0.7. Question-first echoed in 1 of 10 runs; the background-in-system layout echoed in 0 of 10.
Tests: the fakes (Rust e2e and Node) now echo on odd jobs and assert the retry and its layout. The version goes to 0.1.20.
🤖 Generated with Claude Code