FixGraph

Structured troubleshooting for humans and AI agents.

FixGraph provides verified fixes for technical issues and allows AI agents and developers to retrieve structured solutions instead of scraping forums.

Installation

Install globally:

npm install -g fixgraph

Or locally:

npm install fixgraph

CLI Usage

fixgraph "docker container exits immediately"

Example output:

FixGraph CLI
Structured troubleshooting for humans and AI agents
Created by Jawdat Nasar  •  https://fixgraph.netlify.app

🔍 FixGraph Search: "docker container exits immediately"

Docker container exits immediately
Confidence: 93%  |  Verifications: 41  |  Freshness: fresh
  ID: cmmqxav040bfzh1o84imwdxc8

Run with --fix <id> to get the canonical fix.

Commands

fixgraph "<error message>"        # Search for matching issues
fixgraph --fix <issue_id>         # Get canonical fix with steps
fixgraph --id <issue_id>          # Get full issue details
fixgraph --help                   # Show help

JavaScript SDK

import { FixGraphClient } from 'fixgraph'

const client = new FixGraphClient({ apiKey: process.env.FIXGRAPH_API_KEY })

const results = await client.searchIssue('redis ECONNREFUSED 6379')
const fix = await client.getFix(results.items[0].id)
console.log(fix)

Why FixGraph

Traditional troubleshooting resources are designed for humans reading HTML. FixGraph creates a structured troubleshooting graph queryable by:

• Developers at the terminal
• AI coding agents (LangChain, CrewAI, Claude)
• Automation and CI/CD systems
• Debugging workflows

API

GET https://fixgraph.netlify.app/api/search?query=<error>

Response includes: canonical_fix, trust_score, verification_count, compatibility

MCP Server (Claude Code / Cursor)

claude mcp add fixgraph -- npx fixgraph-mcp

Created by

Jawdat Nasar — https://fixgraph.netlify.app