Leverages knowledge graphs to provide compressed context and task orchestration for complex edits.

Install

mkdir -p .claude/skills/graphflow && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/18944" && unzip -o skill.zip -d .claude/skills/graphflow && rm skill.zip

Installs to .claude/skills/graphflow

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Graph-based code context compression, task planning, and knowledge graph orchestration (10 MCP tools). Invoke before ANY code question, bug fix, debugging, file reading, Chinese/CJK query, refactor, or multi-step edit — ALWAYS call graphflow_context MCP first when GraphFlow is connected.
288 chars✓ has a “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Advanced

Key capabilities

  • Compress code context for token efficiency
  • Plan multi-step tasks using a knowledge graph
  • Orchestrate knowledge graph operations
  • Index workspace changes incrementally or fully
  • Export and import graph artifacts for team sharing

How it works

The skill uses a persistent MCP server to index the codebase into a knowledge graph, enabling context compression, task planning, and orchestration. It provides tools for querying context, planning tasks, managing the graph, and gaining insights.

Inputs & outputs

You give it
Code question, bug fix, refactor, or multi-step edit
You get back
Compressed context, task decomposition, or graph insights

When to use graphflow

  • Refactoring complex modules
  • Planning multi-step edits
  • Querying codebase knowledge

About graphflow

Uses a persistent MCP server to index the codebase into a graph, enabling intelligent context compression and multi-step task planning.

Graph-based code context compression, task planning, and knowledge graph orchestration (10 MCP tools). Invoke before ANY code question, bug fix, debugging, file reading, Chinese/CJK query, refactor, or multi-step edit — ALWAYS call graphflow_context MCP first when GraphFlow is connected.

When not to use it

  • When the task is simple and does not require extensive context
  • When the user prefers to scan the whole repository recursively

Limitations

  • The skill requires `graphflow_context` MCP to be called first when GraphFlow is connected.
  • It does not support scanning the whole repository recursively before trying GraphFlow.
  • It does not support reading large files before checking GraphFlow anchors.

How it compares

This skill provides a graph-based approach to context management and task planning, offering token-efficient context and structured task decomposition that differs from manually navigating or searching a codebase.

Compared to similar skills

graphflow side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
graphflow (this skill)015dNo flagsAdvanced
nx-generate16moReviewBeginner
agent-sync-coordinator15moReviewAdvanced
aiox-master03moNo flagsAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

Search skills

Search the agent skills registry