skills.rest
takimoto-sketch
takimoto-sketchCOMMUNITY
Explore all skills by this author →
GitHub

client-knowledge

Collect client company information into a multi-layer knowledge base.

Updated Aug 23, 2026
Search Tags:#documentation#onboarding#claude-code#knowledge-base#gdrive#company-research#client-knowledge
One-click install
npx skills add https://github.com/takimoto-sketch/medica-agent --skill client-knowledge
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: client-knowledge
Source: https://github.com/takimoto-sketch/medica-agent/tree/main/.claude/skills/client-knowledge
Command: npx skills add https://github.com/takimoto-sketch/medica-agent --skill client-knowledge

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Consolidates scattered public research and shared documents into a consistent, session-ready client knowledge base so agents and teams can quickly understand project context without manual aggregation.

Core Features & Use Cases

  • 3-layer knowledge model: Integrates CLAUDE.md references, an operational {client}.md, and detailed research files under knowledge/{client}/research/.
  • Document ingestion and summarization: Reads shared Drive/docs, extracts facts, distinguishes hypotheses from facts, and records document IDs for traceability.
  • Onboarding and continuity: Ideal for client onboarding, project handovers, and ongoing account management where up-to-date context must be preserved across sessions.

Quick Start

Create a complete three-layer client knowledge base for Acme Corp by researching public sources and integrating any provided Drive documents into CLAUDE.md, Acme.md, and knowledge/acme/research/ files.

Frequently Asked Questions about client-knowledge

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

FAQPage Schema
How do I consolidate company research and shared documents into a client knowledge base?▼

To build a client knowledge base, you consolidate web research and shared Drive documents into a 3-layer structure: CLAUDE.md references, an operational {client}.md, and detailed research files under knowledge/{client}/research/.

How does the 3-layer client knowledge model work for onboarding?▼

The 3-layer model works by separating context into CLAUDE.md references for agents, an operational {client}.md for running workflows, and detailed research files for deep background, ensuring project continuity during onboarding and handovers.

Can I ingest shared Drive documents and track where the client information came from?▼

Yes, you can ingest shared Drive documents. The system extracts facts, distinguishes hypotheses from verified facts, and records document IDs under knowledge/{client}/research/ to preserve source attribution and traceability.

What is the best way to maintain client context for ongoing account management across sessions?▼

The best way to maintain client context is generating structured CLAUDE.md references and {client}.md templates that store up-to-date operational details, allowing agents to quickly understand project context without manual aggregation.

Does this approach separate verified facts from hypotheses when researching a client company?▼

Yes, the approach explicitly separates facts from hypotheses. When collecting and structuring client company information from public sources and shared documents, it preserves this distinction within the generated research files.

Related Skills

Explore More Marketing & Sales Skills→
rasmialkharshanrasmialkharshan
2

notebooklm

Integrate Google NotebookLM into Claude Code for source-grounded document queries.

Community
Intermediate
guccangguccang
3

google-workspace

Automate Google Workspace operations across Gmail, Calendar, Drive, Docs, Sheets, and Contacts.

Community
Advanced
luanmorenommacielluanmorenommaciel
231

agent-router

Route data engineering tasks to specialized agents using file patterns and intent keywords.

Community
Advanced
selcukyucelselcukyucel
7

document

Generate CLAUDE.md context files from a module's source code.

Community
Intermediate