deepresearh-integrator_skill

Overview

This skill consolidates multiple deepresearch result files into a single, comprehensive report using an iterative, supervised workflow. It focuses on one file at a time, creates and maintains a TODO plan, and waits for user approval before making changes. The goal is a clean, well-structured final report with preserved attributions and clear traceability.

How this skill works

The skill first scans the workspace and lists all source files and a target consolidated report. It creates a detailed TODO list (one task per source file plus review steps) and presents the plan for user approval. For each approved file, it reads and analyzes content, proposes specific integrations (new sections, updates, conflict handling), and applies changes only after approval, moving processed files to a processed/ directory. Finally, it performs a full review, generates a summary and basic statistics, and ensures consistent formatting and sources.

When to use it

• You have multiple deepresearch output files that must be merged into one authoritative report.
• You need a controlled, auditable integration process with user approval at each step.
• You want to preserve source attribution and handle conflicting findings transparently.
• You require a repeatable workflow that processes files one-by-one to reduce errors.
• You need a final document with clear structure: executive summary, findings, sources, and appendix.

Best practices

• Process files one at a time and keep a TODO list for visibility and recovery.
• Present proposed changes before editing the main report—do not assume consent.
• Group related findings under common headings and preserve source attributions for each item.
• Handle conflicts by reporting both perspectives and flagging unresolved issues for follow-up.
• Keep a processed/ directory and optionally a sources/ directory to maintain backups and traceability.

Example use cases

• Consolidate research outputs from a multi-author project into final-report.md for stakeholders.
• Merge periodic deepresearch exports into a single living document with clear change history.
• Integrate heterogeneous findings (data, textual insights, citations) into themed sections for a literature review.
• Create an executive summary and appendix from several technical deepresearch files for executive briefings.
• Produce a review-ready consolidated report while preserving original files in processed/ for audit.