Observe AI agents that work in GitHub.
Your agent reads repos, opens PRs, comments on issues, dispatches workflows. Trefur captures each GitHub call as a span on the agent trace — without changing the integration code you already wrote.
What you see.
Why it matters.
For engineers
Each GitHub API call your agent makes becomes a tool span. When an agent opens the wrong PR or leaves a review on the wrong file, you see it on the trace before the team does.
For platform teams
Roll up GitHub calls per agent, per repo, per installation. Spot agents writing outside their declared scope. Audit which app identities are most active.
For business operators
Know what your AI is doing in your code — opening PRs, triaging issues, suggesting fixes. Tie agent activity to merge rate, issue close rate, or release cadence.
For SRE
GitHub rate-limit and abuse-detection errors land on the span. Envelope alerts catch retry-rate spikes per agent. Oncall opens the trace with the failing call already highlighted.
How it works.
Your agent calls the GitHub REST or GraphQL API. The Trefur SDK wraps the client and emits a tool span per call — repos.get, pulls.create, issues.createComment, actions.createWorkflow- Dispatch, and the rest. Spans land on the agent trace alongside the LLM call that decided to act, so the full sequence is one navigable timeline. The collector can run inside your network if you want diff or comment content to stay local.
See what your GitHub agents are actually doing.
Free tier. No card required. First trace in under five minutes.