Factory Private
Enterprise AI
Private AI package work needs durable context
September 24, 2026 - 2 minute read
Private AI package maintenance requires an agent to remember more than the dependency being changed. Build conventions, upstream exceptions, reviewer decisions, and failed approaches can span several repositories. Losing those details makes the next package look like a new investigation.
Chainguard’s published Factory story describes one Droid session running for two weeks across six repositories and building 80 packages. Staff Engineer Josh Wolf says:
“When you don't have to think about context windows, you can treat Droid like a colleague that just remembers what you've been talking about.”
That is an account of a customer workflow. It does not establish a universal throughput benchmark, prove that the session was airgapped, or show that every package required the same work.
Private AI package maintenance spans repeated decisions
A package update may start with an upstream change, move through build repair, and end with reviewer feedback that should apply to later packages. The next task needs the conclusion and its rationale. Replaying a long transcript or rediscovering the rule consumes both engineering time and model context.
Factory’s context-compression research describes structured summaries that retain intent, file changes, decisions, and next steps. The published evaluation also identifies file tracking as a weakness across the approaches tested. That limitation is a reason to inspect the working tree and test output, even when the session appears to remember the discussion.
For a pilot, choose a small set of related packages with existing build and review requirements. Preserve the upstream revisions, internal changes, and validation commands. Ask the agent to apply a reviewed convention to a later package, then check whether it preserved the exception that made the first change safe.
Measure repeated investigation and reviewer correction alongside completed packages. Raw package counts can hide differences in complexity, failed attempts, and manual repair.
Private AI package maintenance needs local dependencies
A disconnected runtime still needs source archives, base images, package indexes, signing tools, and test fixtures. Approve how those arrive before giving the agent a maintenance task. A build that silently downloads from a public registry defeats the intended isolation.
Factory’s airgapped deployment documentation describes a runtime without Factory cloud dependencies. Model inference and telemetry must also remain inside the environment. Existing artifact-import procedures remain part of the customer’s operating work.
Keep the agent’s authority narrower than the package-release process. Factory’s Agent Safety & Controls documents command controls and sandboxing. Use those with separate credentials for development, signing, and publication. Preparing a valid patch should not automatically grant permission to publish a package.
The pilot succeeds when another engineer can reproduce the build, inspect the change, and understand which previous decisions were reused. Chainguard’s experience provides a concrete reason to test continuity over sustained work. The private deployment adds a separate requirement to keep that workflow, including its dependencies and evidence, inside the approved boundary.
