Problem statement

OpenSRE currently lacks a first-class incident.io integration. That leaves a gap for teams that coordinate incidents in incident.io and want the agent to consume incident context and publish findings back into the system of record.

Proposed solution

Add a first-class incident.io integration that supports:

• onboarding, authentication, and connectivity verification
• reading incident context, timeline, and metadata relevant to RCA
• incorporating incident.io context into investigation planning and evidence gathering
• optionally posting summaries, findings, or next steps back to the incident when safe and appropriate

Acceptance criteria

• the integration is usable from the investigation workflow end to end
• setup and usage are documented clearly
• the implementation is backed by end-to-end tests in tests/e2e or equivalent
• the PR includes a screen video showing setup plus a successful investigation flow using incident.io
• vibecoding without testing will not be allowed; PRs without working end-to-end coverage and screen-video proof should be rejected