Skip to content

Parse non-JSON AutoResearch strategy plans - #214

Merged
FluffyAIcode merged 1 commit into
mainfrom
AgentMemory/autoresearch-prose-strategy-parser-0720
Jul 20, 2026
Merged

Parse non-JSON AutoResearch strategy plans#214
FluffyAIcode merged 1 commit into
mainfrom
AgentMemory/autoresearch-prose-strategy-parser-0720

Conversation

@FluffyAIcode

Copy link
Copy Markdown
Owner

Summary

  • scan for valid JSON objects without mistaking LaTeX braces for JSON
  • accept fenced Python-style candidate dictionaries
  • convert structured natural-language plans into deterministic candidate inputs

Test plan

  • targeted AutoResearch tests: 39 passed
  • full pytest: 1698 passed, 60 failed, 2 skipped (existing Qwen3/Transformers runtime incompatibility)
  • MLX coverage gate: 1555 passed, 28 failed, 2 skipped; 66.92% coverage (gate fails at required 100%)

The uncommitted candidate.py file is runtime state and is excluded.

Made with Cursor

Ignore LaTeX braces and convert structured natural-language plans into validated candidates so non-JSON Gemma output cannot stop the research loop.

Co-authored-by: Cursor <cursoragent@cursor.com>
@cursor

cursor Bot commented Jul 20, 2026

Copy link
Copy Markdown

Bugbot is not enabled for your account, so this pull request was not reviewed.

Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs.

@FluffyAIcode
FluffyAIcode merged commit 7629d86 into main Jul 20, 2026
7 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant