Factory Droid illustrates a concrete choice for software teams: whether the documented integration, authorization and release lifecycle are compatible with the work they want to delegate. This guide distinguishes supported facts from unknown account, region and permission details as of 25 October 2026.
Product and deployment surface
Factory’s Droid is a code-oriented agent that can be launched interactively or in a noninteractive command-runner form. The official CLI reference distinguishes inspecting code from writing files and running more consequential commands. This makes it a practical case study in why agent risk cannot be reduced to one “auto” toggle: the permission policy, execution surface, network scope and session mode jointly determine what happens.
Worked engineering scenario
For a concrete trial, start by asking the CLI to explain an unfamiliar test failure with default read-only access. Next, isolate a branch and permit a focused edit. Only then allow package installation and test execution under a documented policy. The CLI’s `rules check` procedure helps inspect a command decision without executing that command, but a policy preview should not be mistaken for a full runtime safety proof.
Repository authority and data exposure
Documentation separates low-risk file operations from development commands such as dependency installation and from higher-risk Git pushes. Headless execution is especially sensitive: it cannot pause for every interactive question, and a script may interpret exit codes or outputs incorrectly. Keep dangerous permissions out of ordinary developer shells and use an ephemeral environment for experiments with elevated access.
Version, licensing and operational limitations
The enterprise security documentation describes local file actions, authorization and rule enforcement, but prompt context and selected code may still be sent according to model and gateway settings. A local CLI therefore does not imply that private source never crosses a network. Check provider routing, secret filters, network egress and retention against the contracted deployment.
When the product fits and when to wait
Factory Droid is useful for teams that can establish a permission-as-code policy and need both a human terminal workflow and reproducible automation. Record installed CLI version, policy version, exact model selection, user identity and reviewer sign-off for each evaluated change.
Alternatives and selection trade-offs
Compare Codex CLI, Claude Code, Goose and OpenCode under the same bounded repository task. A high-level product promise of autonomous delivery is not a substitute for diff review, CI, signing and merge protections.
A reproducible evaluation protocol
- Use a disposable, nonproduction repository at a recorded commit and remove real access tokens, private keys, client information and production data.
- Write acceptance criteria before engaging the assistant: at least one deterministic test, a static check and one failure or rollback scenario.
- Capture the actual tool version, model/provider route, authorized repository, permitted commands and network scope rather than relying on product branding.
- Inspect every changed file and generated dependency. Re-run the build, unit tests and security checks outside the assistant’s self-reported output.
- Separate evidence: vendor documentation, observed measurements from a future controlled trial, and editorial assumptions must have distinct labels.
What was verified—and what was not
The primary evidence for this article is vendor documentation or an original project repository available by the 25 October 2026 editorial cutoff. A product description is evidence of how its publisher represents a capability, not an independent security audit or benchmark. The exact installed build, account entitlement, country, processor, contract and provider model may change the result. We did not install or benchmark the tool for this article.
Availability, licensing and pricing boundary
Snapshot dated 25 October 2026. The verified material here does not provide a complete current price, seat allocation, model quota, region list, retention policy or binding commercial license for each possible tenant. Those entries are deliberately null or unresolved in the structured record. Consult an official account-specific agreement, and link any future numeric plan snapshot to the canonical AI10 pricing registry rather than reproducing a stale marketing offer.