Skip to content
Integrations

Integrations

Everything RunLore plugs into: what can trigger an investigation, which LLMs run it, the data sources it reads for signal, where notifications land, and the Git forge it writes findings back to. Each page below is short by design — the minimal config to wire the integration, how to verify it locally, and the gotchas that would otherwise cost you a debugging session.

Presence is enablement. A key under sources.<name> — or the equivalent notifier / data-source block — is what turns that integration on; there is no separate enabled: true flag to hunt for. Leave a data source out and RunLore simply runs without the tool it would have unlocked: no source, notifier, or data source is required, and none is assumed. See Configuration for the exhaustive key reference every page here deep-links into.

Triggers

What starts an investigation.

LLM providers

What runs the investigation loop.

Data sources

What the agent reads for signal — every one is optional; an unset data source just disables the tool it would have unlocked.

Notifications

Where findings are delivered.

Forge

Where the Learn loop writes findings back to.