FaultLine Wiki

Main page

This collection covers how to set up and run FaultLine — connecting an AI, how memory is written and corrected, seats and configuration, endpoint health and troubleshooting, billing — together with reference notes on the benchmark we measure memory against.

Handbook

Thirty-one topics, hosted in the product console.^[1]

Starting out

• Quick start
• Home
• What FaultLine is
• Connecting your AI
• Connecting to FaultLine

How memory works

• Memory: how FaultLine remembers
• Correcting a memory
• Growing the memory
• The agent's own memory

Connecting your tools

• MCP, explained
• Developer API
• Unencrypted (http://) endpoints

Administration

• The consoles
• Managing seats & logins
• Configuration
• Privacy & your data
• Clinical / regulated lane

Health & troubleshooting

• Troubleshooting: where to start
• Chat-agent troubleshooting
• My endpoint isn't working
• Endpoint state: FAIL
• Endpoint state: FLAPPING
• Endpoint state: DEGRADED
• Latency profiles & calibration

Chat agent

• The FaultLine chat agent
• Chat-agent trial & subscribe

Plans & usage

• Pricing & tiers
• Memory & usage

Reference

• Glossary
• FAQ & troubleshooting
• Roadmap

Measurement

• LongMemEval — the long-term memory benchmark, described from its primary sources.^[2]

References

1. FaultLine Help Wiki — app.faultline.ca/dashboard/help/
2. Di Wu, Hongwei Wang, Wenhao Yu, Yuwei Zhang, Kai-Wei Chang, Dong Yu. LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory. ICLR 2025. arXiv:2410.10813 — arxiv.org/abs/2410.10813

Last updated: 2026-08-10