Mini Context Graph
Install mini-context-graph when you want a lightweight way to structure and reuse agent context as a graph instead of ad-hoc chat history.
Overview
mini-context-graph is an agent skill most often used in Build (also Ship, Operate) that helps solo builders maintain a compact context graph for coding agents instead of unstructured chat logs.
Install
npx skills add https://github.com/fizznix/mini-context-graph --skill skill.mdWhat is this skill?
- Targets mini context-graph workflows for solo builders shipping with agent-first editors
- Complements procedural memory and RAG by emphasizing linked entities rather than flat prompts
- Fits local or repo-adjacent agent setups where you control context structure
- Useful when sessions grow long and you need durable context nodes instead of re-explaining the codebase
- SKILL.md body in the ingest snapshot is minimal—verify commands and schema in the repo before relying on it in productio
Adoption & trust: 1 GitHub stars.
What problem does it solve?
Your agent keeps losing thread across sessions because project knowledge lives only in ephemeral messages, not in a small linked context model you can reuse.
Who is it for?
Solo builders experimenting with graph-shaped agent context alongside Claude Code, Cursor, or Codex when SKILL.md in the repo documents the actual workflow.
Skip if: Teams that need a fully documented, enterprise graph platform in the catalog entry alone—the ingested SKILL.md here is essentially empty.
When should I use this skill?
When you need a compact, graph-shaped context layer for an AI coding agent and the repository SKILL.md defines how to create and query it.
What do I get? / Deliverables
You get a repeatable context-graph habit so agents can anchor work to named nodes and relationships rather than re-deriving the whole repo every time.
- A defined mini context-graph schema or files the agent maintains
- Linked context nodes usable across coding sessions
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Catalog listing places it on Build because context graphs are created while wiring agent workflows and tooling, even though the graph is consulted again in Ship and Operate. Agent-tooling is the canonical shelf for skills that shape how coding agents hold, link, and retrieve project context.
Where it fits
You model repo areas as graph nodes so the agent pulls related files and decisions in one pass.
You attach review findings to existing context nodes instead of opening a new unstructured thread.
You trace a production fix back through linked decisions the graph captured during the original build.
How it compares
A lightweight agent-skill pattern for structured context links, not a hosted knowledge-base MCP server or full procedural-memory CLI.
Common Questions / FAQ
Who is mini-context-graph for?
Solo and indie builders who run AI coding agents daily and want a minimal context-graph approach before investing in heavier memory or RAG stacks.
When should I use mini-context-graph?
During Build when you define how agents store entities and edges; during Ship when you want review context tied to decisions; during Operate when incidents need the same linked project map—after you confirm the repo SKILL.md matches your workflow.
Is mini-context-graph safe to install?
Review the Security Audits panel on this Prism page and read the full skill package in the repository before granting filesystem or shell access to your agent.
SKILL.md
READMESKILL.md - Mini Context Graph
| # mini-context-graph { "name": "mini-context-graph", "description": "|" }