Code graph + runtime log intelligence for engineering teams

Understand code, traces, and logs
before AI touches production

Log Thread maps your repository locally, connects to your existing observability stack, and correlates runtime failures back to the exact files, functions, dependencies, and downstream impact. Your team gets a fast local Web UI, CLI workflows, and optional MCP-powered IDE assistance.

Works in the Log Thread Web UI and CLI, and also plugs into Cursor, Windsurf, Claude Code, and other MCP-compatible tools.

Local workspace graph

Issue: Checkout latency spiked after deploy

✦ Log Thread: Impact path touches checkout_service.py

✦ Related symbols: payment retry helper, cart serializer, order repository

✦ Downstream impact: billing API, queue worker, order sync job

Next: open the runtime panel to correlate logs and traces

Runtime log analysis

$ logthread logs query prod-loki '{service="checkout"} |= "timeout"'

Events: 148 · Patterns: 5 · Sequences: 2

Top answer: timeout burst maps to reserve_inventory()

Evidence: retries began after dependency call to inventory gateway

Suggested follow-up: inspect retry loop and queue backpressure paths

Install in minutes, connect the data you already have

Start with the local CLI, analyze your repository, then connect runtime providers such as Loki, Datadog, Splunk, Sumo Logic, Elasticsearch, CloudWatch, or New Relic.

1

Install

Run pip install logthread

2

Authenticate & Generate

Use logthread login and logthread generate .

3

Connect Logs

Add a provider with logthread logs connect ... or in the local Web UI

4

Investigate Anywhere

Use logthread ui, the dashboard, CLI analysis, or MCP-assisted editing

Bring your existing runtime data sources

Log Thread is designed to sit on top of your current stack, not replace it. Point it at your existing observability systems and correlate production evidence back to the code graph.

Native log providers

Connect Grafana Loki, Datadog Logs, Splunk, or Sumo Logic with provider-specific credentials and saved defaults.

logthread logs connect --provider loki --name prod

Enterprise observability

Use Elasticsearch, CloudWatch, and New Relic through the same local runtime panel and query workflows.

Great for teams that already have governed logging and security controls in place.

Correlation-first analysis

Attach request IDs, trace IDs, service names, and other hints so Log Thread can cluster failures into actionable root-cause paths.

No need to duplicate logs or move your code into the cloud.

Individual and enterprise workflows are different on purpose

Individual users start with a personal workspace. Enterprise access is organization-based: your admin sets up a shared organization, verifies the company domain, enables SSO, and assigns seats.

Individual users

Create an account, verify your email, subscribe, then run logthread login, logthread generate ., and logthread ui.

Best for single users and small teams getting started quickly.

Enterprise teams

Enterprise access is not a per-user flag. It is a company organization with verified email domains, configured SAML or OIDC, and seat-managed memberships. Team members simply enter their work email at sign-in and Log Thread routes them to the right identity provider.

Use this for Okta and other SAML/OIDC-backed rollouts with centralized access control.

What changes when code and runtime context live together

Root cause faster

Trace production failures from log lines and request sequences back to the functions and modules that actually changed.

Reuse what already exists

Graph context helps engineers and AI assistants find the right helpers, interfaces, and dependency boundaries instead of re-implementing them.

Know what breaks

Impact analysis shows the APIs, workers, tables, and downstream workflows touched by a proposed fix before it ships.

Use the interface you want

Investigate from the local Web UI, run CLI analysis in the terminal, or surface the same context through MCP-enabled editors.

Operationalize install and onboarding

New teammates can get from install to first indexed repository and connected log source in a few guided steps.

Keep data local-first

Code analysis stays on the machine, while cloud features focus on auth, seats, billing, and team access control.

Simple, transparent pricing

One plan for code intelligence, runtime analysis, team auth, and enterprise rollout.

Most popular
$10.99 /user/month
  • Unlimited repositories per user
  • Code graph, trace, dependency, and impact analysis
  • Runtime log analysis with existing providers and correlation hints
  • CLI, local Web UI, hosted dashboard, and optional MCP workflows
  • Enterprise SSO and org-managed seats
Get Started Now