Polygraph is a meta-harness for maximum agentic autonomy. It doesn't replace the agent you already use (Claude Code, Codex, OpenCode), it adds the components agents need to work effectively across a real organization.
Agents hit two walls. In space, an agent is stuck in one repo: it can't see how a change fits the wider system, and it can only write to one repo at a time. In time, it forgets: every session starts from a blank slate, so a human carries the context. Both walls cap autonomy.
Polygraph lifts both. It connects all your repos (private and public) into a unified dependency graph without moving any code, so your agent reads and writes across every repo you have access to and orchestrates the PRs and CI as one change. And it records every agent session, so you can resume, reference, or build on work done by anyone, on any machine, even with a different agent. Think of it as a synthetic monorepo: monorepo ergonomics, no migration.
What's next
Section titled “What's next”Use our Quickstart guide to get started with Polygraph right away.
Or explore our features:
- Cross-repo agent sessions: choose the repos a session should use and expand the working set as the task crosses repo boundaries.
- Cross-repo PR and CI coordination: open, link, track, and verify related PRs across repos from one session.
- OSS repo integration: bring public OSS repos into a session for repros, upstream debugging, and integration validation.
- Memory and resumability: resume a session tomorrow, reference it from a new session, or hand it to a teammate.
Try out Polygraph in our example use cases:
- Ship feature across repos: one feature that spans a backend and its frontend, or a contract across several microservices.
- Ship library change to consumers: validate a design system or library change against every consumer, then PR the clean ones and flag the rest.
- Patch a CVE across every affected repo: find every affected repo, open patch PRs, and hand each to the owning team with the audit trail intact.
- Ask questions across repos: pattern transfer, impact analysis, and investigation, when the read is the deliverable.
- One session for the whole sprint: resume or reference a session instead of re-explaining context at every handoff.
- For OSS developers: triage contributor repros and validate fixes against the consuming repo.