Resources for Factory

A managed model for producing software with repeatable inputs, standardized tooling, and measurable output.

The shift from coding agents to full software factories, and what autonomy across the SDLC actually looks like.

The practices that make a repository legible to coding agents — clear structure, tests to verify against, and written-down conventions.

How Factory defers and compresses context to keep agents effective across long, complex tasks.

How large enterprises automate software development in 2026, with concrete examples drawn from published case studies.

Running the open-weight DeepSeek V4 model natively inside a Droid — selection, reasoning levels, and cost.

Mapping agent usage to delivery speed, work priorities, and shipped artifacts across an engineering organization.

A reference architecture: the components that move a unit of intent through specification, execution, review, and deployment.

Running Zhipu AI's open-weight GLM models in a Droid at a low credit multiplier for high-volume work.

A learned approach to detecting secrets and preventing leaks in agent-generated code.

Turning codebase quality into a score you can track — the CLI report, dashboard, and API behind agent readiness.

Central governance with explicit boundaries for access, integrations, policy, spend, and data residency.

Using Moonshot AI's Kimi K2 family for scoped coding tasks with open weights and a low credit multiplier.

A reference comparison of platforms implementing the software factory model — delivery, audit, model flexibility, and deployment.

A community of students, builders, and researchers shaping how software gets built with Droids.

The case against coupling your workflow to a single model provider, and what switching models per task buys you.

Running MiniMax, the lowest-multiplier models in the catalog, for cost-sensitive batches and cleanup passes.

How Factory Router picks the right model for each step automatically instead of pinning a run to one.