# Mcp Knowledge Graph

> Use when starting work on a codebase you want to remember across sessions, extracting architectural facts, or querying persistent project knowledge instead of re-reading files. Centers the knowledge-graph MCP and project-local `.aim` storage.

- **Type:** Skill
- **Install:** `agentstack add skill-spideynolove-claude-dotfiles-mcp-knowledge-graph`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [spideynolove](https://agentstack.voostack.com/s/spideynolove)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [spideynolove](https://github.com/spideynolove)
- **Source:** https://github.com/spideynolove/claude-dotfiles/tree/main/.agents-global/skills.disabled/mcp-knowledge-graph

## Install

```sh
agentstack add skill-spideynolove-claude-dotfiles-mcp-knowledge-graph
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# MCP Knowledge Graph

## When To Use

Use this skill when:
- Starting work on a codebase you want to remember across sessions
- Extracting and persisting architectural facts like components, dependencies, and decisions
- Building a project knowledge graph from broad codebase analysis
- Querying previously stored architectural context instead of re-reading files

Do not use it for temporary session notes, raw code dumps, or general chat memory.

## Core Workflow

1. Gather broad codebase context first, often with `repomix`
2. Extract durable facts, not raw code
3. Store entities with `aim_memory_store`
4. Store relations with `aim_memory_link`
5. Append facts with `aim_memory_add_facts`
6. Query later with `aim_memory_search` or `aim_memory_get`

## Storage Choice

| Use case | `location` | `context` | Result |
|----------|------------|-----------|--------|
| Team or project knowledge | `"project"` | omit | `.aim/memory.jsonl` in the repo |
| Personal notes for a project | `"global"` | omit | global memory file |
| Separate named personal graph | `"global"` | set | named global memory db |

Default to `location: "project"` with no `context` so the graph is easy to discover and commit.

## Main Operations

| Tool | Purpose |
|------|---------|
| `aim_memory_store` | Create entities |
| `aim_memory_link` | Create typed relations |
| `aim_memory_add_facts` | Append observations |
| `aim_memory_search` | Keyword search across the graph |
| `aim_memory_get` | Exact lookup by entity name |
| `aim_memory_read_all` | Dump a graph for review |
| `aim_memory_list_stores` | Discover available stores |
| `aim_memory_forget` | Delete entities |
| `aim_memory_remove_facts` | Delete observations |
| `aim_memory_unlink` | Delete relations |

## Entity Design

- Use stable entity names like `AuthService`, `UserRepository`, `BillingFlow`
- Use semantic types like `service`, `repository`, `module`, `decision`, `concept`
- Store facts as concise observations
- Use active relation verbs like `depends_on`, `calls`, `owns`, `implements`

## Search Limits

- `aim_memory_search` is substring-based, not semantic
- `aim_memory_get` is exact-name lookup
- Observation wording matters; use predictable terminology

## Example Pattern

1. Store components and decisions
2. Link them with explicit relations
3. Query by keyword before re-reading the repo

Prefer focused graphs over huge ones because writes rewrite the whole file.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [spideynolove](https://github.com/spideynolove)
- **Source:** [spideynolove/claude-dotfiles](https://github.com/spideynolove/claude-dotfiles)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-spideynolove-claude-dotfiles-mcp-knowledge-graph
- Seller: https://agentstack.voostack.com/s/spideynolove
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.