# Ccc Research

> complete research ecosystem — 8 skills in one. Deep multi-source research, spec interviews, cross-model review, competitive analysis, trend analysis, and large…

- Skill: `kevinzai/ccc-research-2` (Agent Skill)
- Install (CLI): `npx skillmds@latest add kevinzai/ccc-research-2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/kevinzai/ccc-research-2/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: KevinZai (https://skillmd.com/u/kevinzai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/kevinzai/ccc-research-2

---


# /ccc-research — Research domain hub

> Load ONE skill. Get the entire research domain. 8 skills in one.

## Sub-Skills

| # | Skill | Focus |
|---|-------|-------|
| 1 | deep-research | Multi-source research with parallel agents, citation tracking, synthesis |
| 2 | spec-interview | 5-7 structured questions to produce a complete spec before coding |
| 3 | cross-model-review | Code/decision review using multiple AI models for diverse perspectives |
| 4 | literature-review | Academic/technical literature review with source evaluation |
| 5 | competitive-analysis | Competing products/tools/libraries feature comparison |
| 6 | data-ingestion | Ingest and summarize large documents, codebases, or datasets |
| 7 | trend-analysis | Technology, market, or usage pattern trend analysis |
| 8 | large-context | Opus 5's native 1M window, reduction, chunk-and-synthesize, then Codex CLI for mechanical execution |

## Research Surfaces

In addition to web search, research can use connected channels as surfaces:
- **Telegram/Discord channels** — community sentiment, ecosystem news, real-time signals
- **Files API** — ingest uploaded documents, codebases, or data files directly
- **Cross-model review** — route to multiple models for independent perspectives

> **No expressed intent?** Present the top 3 sub-skills + "More…" as an AskUserQuestion picker (≤4 options).

## Routing Matrix

| Your Intent | Route To |
|-------------|----------|
| "Research X before building" | `deep-research` + `competitive-analysis` |
| "Define spec first" / "Interview me" | `spec-interview` |
| "Review this from multiple angles" | `cross-model-review` |
| "Literature review" / "Academic sources" | `literature-review` |
| "Compare these tools/libraries" | `competitive-analysis` |
| "Summarize this large document" | `data-ingestion` |
| "What are the trends in X?" | `trend-analysis` |
| "File too large for context" | `large-context` |

## Protocol: Research Before Build

The recommended sequence for any non-trivial feature:
1. `deep-research` → find existing solutions, prior art, best practices
2. `competitive-analysis` → evaluate options
3. `spec-interview` → define requirements before writing code
4. Hand off to implementation skills with research context

## When to invoke this skill

**Example 1**
- user: research the best auth libraries for Next.js before I start building
- assistant: Loads ccc-research and routes to deep-research + competitive-analysis for multi-source comparison with citation tracking.

**Example 2**
- user: help me define the spec for this feature before coding
- assistant: Loads ccc-research and routes to spec-interview — 5-7 structured questions to generate a complete spec before implementation.

**Example 3**
- user: analyze this 500-page PDF of technical documentation
- assistant: Loads ccc-research and routes to data-ingestion for large document summarization and key insight extraction.

---

> ⚙️ **Fable contract:** plan before build · verifier ≠ worker · prove before alarm · loops need gates · leave durable state — `rules/fable-method.md`