mm-report-reviewer

Automating quality assurance for research reports requires scoring drafts across multiple dimensions like factuality and evidence coverage. This skill evaluates reports to determine pass/fail outcomes and generates targeted revision briefs to guide analysts in fixing identified issues.

Multi-dimensional scoring validates research reports by evaluating factuality, evidence coverage, decision quality, market context, company specificity, and risk discipline. This mechanism scores each dimension against configured thresholds to produce an objective pass/fail decision and highlight specific blockers.

Generating a revision brief involves submitting the draft review packet for assessment. When reports fall below quality thresholds, the system outputs structured JSON containing dimension scores, identified blockers, and actionable guidance to direct writers toward fixing specific deficiencies.

MCP workflow integration supports automated research report reviewing by producing a structured review output compatible with MCP-driven pipelines. This allows autonomous research systems to archive review history and objectively validate analyst drafts without manual intervention.

Autonomous research pipelines can validate analyst drafts by applying objective quality checks and evidence coverage validation. This skill fits into such pipelines by scoring drafts across six dimensions, ensuring reports meet factuality and risk discipline standards before approval.

Automated report quality scoring relies on resolved configuration thresholds to determine pass/fail outcomes, meaning its effectiveness depends on the accuracy of predefined standards. It outputs structured JSON with scores and blockers, requiring human interpretation to execute complex revision strategies.