# superloglabs/superlog

Open-source observability tool that uses AI agents to self-heal your software

Repository: https://github.com/superloglabs/superlog
Canonical: https://ross.abutalabs.com/products/superlog
Homepage: https://superlog.sh
Language: TypeScript
License: Apache-2.0
License Family: permissive
Topics: observability, opentelemetry, ai, llm, memory, skills, react, typescript, self-hosted
Last push: 2026-08-26T04:32:58+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 6
- inputs: {"age_days": 92, "days_push": 7, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1402, forks 105 (observed 2026-08-28T04:04:37.407939+00:00)

## What it is
Superlog is an open-source agentic observability platform that ingests OpenTelemetry traces, logs, and metrics, groups noisy error signals into incidents, and dispatches AI agents to investigate root causes and open fix PRs automatically. It ships as a self-hostable community edition (web app, OTLP ingest proxy, workers, Postgres/ClickHouse backend) plus a hosted cloud offering with Datadog/Sentry imports and Slack integration.

## Use cases
- self-host an OpenTelemetry observability stack for traces logs and metrics
- group noisy errors into incidents automatically
- have an AI agent investigate production errors and open fix PRs
- import errors from Datadog or Sentry into an open-source workspace
- query high-cardinality telemetry locally with ClickHouse
- get incident context in my coding agent via MCP
- monitor Vercel Railway Render AWS or Cloudflare logs without manual wiring
- ask questions about my telemetry in Slack

## When to choose
- you want an open-source, self-hosted alternative to Datadog or Sentry with OpenTelemetry-first ingestion
- your team wants automated root-cause investigation and AI-generated fix PRs from production errors
- you need a local-first telemetry workspace with incident grouping and ClickHouse-backed queries
- you want incident context available to coding agents through an MCP server

## When to avoid
- you need a battle-tested, long-proven APM for large enterprise fleets with strict compliance requirements
- you only need simple log aggregation without AI investigation features
- you cannot run Docker/Postgres/ClickHouse infrastructure or prefer a fully managed vendor
- your telemetry stack is not OpenTelemetry-based and you lack Datadog/Sentry to import from

## Facets
- artifact type: service
- maturity: active
- function: monitoring, tracing, logging, alerting, agent-framework, mcp, llm-inference
- domain: monitoring, artificial-intelligence, developer-tools, self-hosted
- platform: self-hosted, cloud
- tags: observability, opentelemetry, agentic-observability, incident-management, self-healing, otlp, clickhouse, fix-prs, datadog, sentry, devops, ai-agents, docker, nodejs, web-server

## Member repositories
- superloglabs/superlog (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:37.407939+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:38:59.899619+00:00, confidence not recorded.
  - readme: https://github.com/superloglabs/superlog (fetched 2026-08-28T04:04:37.407939+00:00, sha 15d4301ed633)
  - homepage: https://superlog.sh (fetched 2026-08-29T11:53:11.339478+00:00, sha ceb17a9c830a)
  - site_page: https://docs.superlog.sh (fetched 2026-08-29T11:53:11.342254+00:00, sha 5a91b0d9c34c)
  - site_page: https://superlog.sh/changelog (fetched 2026-08-29T11:53:11.343951+00:00, sha 8b6256ff23b3)
  - site_page: https://superlog.sh/pricing (fetched 2026-08-29T11:53:11.346045+00:00, sha 9bde9405aaca)
- Data as of 2026-08-30T08:39:29.467469+00:00.