combine

Synthesizes prior local research, task artifacts, and git history into a unified report.

2|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/DoctorMozg/claude-pipelines --skill combine-doctormozg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: combine
Source: https://github.com/DoctorMozg/claude-pipelines/tree/main/plugins/mz-research-pipe/skills/combine
Command: npx skills add https://github.com/DoctorMozg/claude-pipelines --skill combine-doctormozg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Scattered pipeline output — past research reports, task artifacts, reviews, and git history — is hard to consolidate by hand. This Skill harvests those local sources and folds them into a single unified report with cross-referenced findings, conflicts, and residual gaps. ## Core Features & Use Cases - Local-first synthesis: Inventories .mz/research/, .mz/task/, .mz/reports/, .mz/reviews/, codebase files, and git history, then decomposes the task into 3–6 parallel research lenses. - Approval-gated pipeline: Presents the source inventory and any residual gaps for explicit user approval before dispatching agents or optional web gap-fill research. - Task-adaptive reports: Generates report sections from the task text (or a user-supplied sections: list), always appending Timeline, Conflicts, Gaps, Sources, and Methodology meta-sections. - Use Case: After several weeks of investigation, run /combine synthesize our findings on the auth refactor to merge prior deep-research reports, task artifacts, and commit history into one cited report. ## Quick Start Ask the AI to combine all prior research and task artifacts about a specific topic into a single consolidated report.

Frequently Asked Questions about combine

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I consolidate prior research and task artifacts into one report?

Invoke the combine skill with a task description, optionally adding output, sources, scope, or sections parameters. It inventories local sources, dispatches parallel lens agents to extract findings, and writes a unified report under .mz/reports/.

What is the difference between combine and deep-research?

Combine is local-first and assumes prior pipeline output already exists on disk, consulting the web only to fill residual gaps behind an approval gate. Deep-research is web-first and starts from zero when no local research exists.

Can I control which sections appear in the combined report?

Yes, pass a sections parameter with a comma-separated list and those entries become the top-level report headings verbatim, in the order given. Without it, sections are derived from noun phrases in the task text, and meta-sections like Gaps and Sources are always appended.

Does combine work in a directory that is not a git repository?

Yes. If git commands fail, the git_history lens is marked unavailable, recorded in state, and skipped without counting toward the three-lens floor. The remaining local sources are still synthesized, and the degradation is surfaced at the approval gate.

When should I not use the combine skill?

Avoid it when no prior local research exists (use deep-research instead), when you want to understand how code works (use explain), or when you need a fixed-template compliance report. Combine is for synthesizing existing on-disk knowledge.