review-research-report

What problem does it solve?

open-xquant's deterministic quality assurance steps only verify file integrity, hash correctness, and basic formatting of research reports. Subtle semantic flaws like decision policy mismatches, misrepresented audit findings, unsubstantiated risk claims, and inconsistent chart narratives can slip through these checks, leading to unreliable strategy promotion decisions.

Core Features & Use Cases

• Decision Consistency Validation: Verifies that the report's final strategy decision aligns with the declared decision policy, audit results, and robustness test outcomes, blocking promotion of strategies that fail declared constraints.
• Audit & Robustness Fidelity Review: Confirms that all audit findings, reproducibility status, and robustness test results are accurately represented in the report, preventing vague language like "risk is controllable" from glossing over material warnings.
• Numeric Claim Triage: Classifies unverified numeric warnings to separate actual report errors from facts registry gaps or low-impact appendix issues, ensuring critical metric mismatches are not missed.
• Chart & Semantic Consistency Check: Validates that charts are readable, correctly ordered per the canonical report structure, and that the HTML version of the report matches the Markdown source without meaning changes. Use case: Quant researchers and AI agents building automated quant research workflows can use this skill to catch report flaws before finalizing strategy decisions, avoiding the promotion of strategies with hidden performance flaws or misrepresented risk profiles.

Quick Start

Use the review-research-report skill to audit the research report for run_id 20240520-001 after all deterministic QA steps have completed successfully.