microsoft/OpenRCA

[ICLR'25] OpenRCA: Can Large Language Models Locate the Root Cause of Software Failures?

53
/ 100
Established

Comprises a benchmark dataset across three production systems (Telecom, Bank, Market) with multi-modal telemetry including KPI time series, dependency traces, and semi-structured logs. Introduces RCA-agent, a Python-based agentic baseline that retrieves and analyzes telemetry programmatically rather than processing full contexts, reducing token overhead while maintaining reasoning capability. Supports custom task generation and evaluation with standardized metrics for assessing LLM performance on root cause localization in complex distributed systems.

292 stars.

No Package No Dependents
Maintenance 10 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 17 / 25

How are scores calculated?

Stars

292

Forks

36

Language

Python

License

MIT

Last pushed

Feb 24, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/microsoft/OpenRCA"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.