deep-research

Orchestrate an 8-phase Python research pipeline with credibility scoring.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/junkijin/my-opencode --skill deep-research-junkijin
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deep-research
Source: https://github.com/junkijin/my-opencode/tree/main/skills/deep-research
Command: npx skills add https://github.com/junkijin/my-opencode --skill deep-research-junkijin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill automates deep, citation-backed research by orchestrating multi-source collection, validation, and synthesis to deliver credible reports, reducing manual hours and cognitive load.

Core Features & Use Cases

  • 8-phase pipeline: Scope → Plan → Retrieve → Triangulate → Synthesize → Critique → Refine → Package with built-in validation and credibility scoring.
  • Source credibility scoring: Quantifies each citation from 0-100 and flags potential biases.
  • Autonomous operation: Default to plan and execute with minimal user prompts; optional graceful escalation when needed.
  • Progressive context management: Caches static context and loads dynamic content on demand to optimize latency.
  • Multiple depth modes: Quick, Standard, Deep, and UltraDeep to balance speed and thoroughness.
  • Self-contained, offline-friendly: Pure Python stdlib usage with deterministic outputs and local storage for reports.
  • Structured outputs: Markdown reports with Bibliography, Methodology Appendix, and validation status; HTML/PDF variants available.
  • Extensive validation: Automated report validation with 8 checks and a target of 10+ sources per report.

Quick Start

  • Provide a concise instruction to run the quick-start: Use the quick start or CLI to run the module with a sample query.

Frequently Asked Questions about deep-research

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

FAQPage Schema
How do I generate a citation-backed research report autonomously?▼

To generate a citation-backed research report autonomously, provide a complex topic query and the skill runs an 8-phase pipeline to gather, validate, and synthesize sources into a Markdown report. It defaults to planning and executing with minimal user prompts.

What is source credibility scoring in automated research?▼

Source credibility scoring in automated research quantifies each citation from 0-100 to evaluate reliability and flags potential biases. This validation ensures synthesized reports maintain rigorous verification standards across all gathered sources.

Can I run deep research workflows offline using only Python?▼

You can run deep research workflows offline using pure Python standard library modules. The self-contained script operates deterministically without external dependencies and stores generated reports locally.

What is the best way to handle complex topics requiring 10+ sources?▼

The best way to handle complex topics requiring 10+ sources is using the Deep or UltraDeep depth modes. These modes balance speed and thoroughness while orchestrating multi-source triangulation and synthesis for rigorous verification.

Does deep research output include a methodology appendix and bibliography?▼

Deep research output includes structured Markdown reports with a bibliography, methodology appendix, and validation status. HTML and PDF variants are also available for export.

Why does autonomous research use progressive context loading?▼

Autonomous research uses progressive context loading to optimize latency by caching static context and loading dynamic content on demand. This ensures efficient processing during the retrieval and synthesis phases.