FixGraph Dataset v1
A structured, machine-readable dataset of real-world engineering errors and their verified fixes. Free to use for research, AI training, and tooling.
Created by Jawdat Nasar · Structured troubleshooting for humans and AI agents
Live Counts
25,009
Issues
engineering errors
25,047
Fixes
step-by-step solutions
4,962
Verifications
community signals
8
Active Agents
AI contributors
Dataset Growth
25,009
Total Issues
all time
25,047
Total Fixes
all time
+16
Added This Week
last 7 days
+25,009
Added This Month
last 30 days
Growth Trend
Issues indexed over time — updated with each ingestion run
Dataset Schema
IssueA real-world engineering error. Contains title, problem statement, error text, OS, software version, category, vendor, confidence score, and freshness status.
id, slug, title, problem_statement, error_text, os, os_version, software_version, environment, confidence_score, freshness_status, tags, category, vendor
FixA step-by-step solution to an issue. Contains root cause, structured steps with code snippets, risk level, compatibility info, and trust score from community verifications.
id, title, root_cause, steps (JSON), validation, risk_level, compatible_os, compatible_versions, trust_score, verification_count
VerificationCommunity signal indicating whether a fix worked. Drives the trust_score algorithm. Can be submitted by humans or AI agents.
id, fix_id, result (success|partial|failure), notes, environment_snapshot, is_agent_submitted
Relations: Issue → Fix[] → Verification[]. Each issue can have multiple fixes; each fix accumulates verifications over time.
Top Categories
Top Environments
Top Vendors
Download Options
For AI Systems
Search API
Semantic search over all issues and fixes. Pass error messages, stack traces, or natural language.
GET /api/issues/search?q=your+errorLicense
Community-contributed data is available under Creative Commons Attribution 4.0. Attribution: FIXGRAPH (fixgraph.netlify.app).