FAULTLINE builds open-source tools for AI agent reliability: measure repeated runs, check recorded traces before a release, and monitor MCP tool calls at runtime. Explore the platform.
| Product | Use it for | Install | Details |
|---|---|---|---|
| faultgate | Check recorded AI agent traces for failure modes and business-rule violations; compare candidate runs with measured baseline variation and return PASS, FAIL, or INCONCLUSIVE in CI. | pip install faultgate |
Source · PyPI · Docs |
| MCP Guard | Monitor local stdio MCP servers for changed tool definitions and tool-description poisoning; audit calls by default and block supported rule violations in enforce mode. | pip install mcp-integrity |
Source · PyPI |
Naming: FAULTLINE MCP Guard is the product. Its Python package and command are named mcp-integrity; install it with pip install mcp-integrity.
The FAULTLINE research workbench contains reproducible studies of agent failures and reliability. Read the research pages for methods and results.
Built and maintained by Samir Sawarkar (@samirsawarkar). Website · Contact