Forge SDLC blog

Short articles and updates about Forge SDLC as a product and operating model. For the reusable engineering framework, see the Blueprints handbook blog and the handbook home.

AI-native delivery series

Evidence-led articles on AI-era software delivery—recommended reading order:

1. The New Bottleneck Is Verification, Not Coding
2. AI Value Comes From Workflow Redesign, Not Tool Rollout
3. Governance Is Becoming a Performance Function
4. Velocity Is a Weaker Story in the AI Era
5. Vibe Coding Changes the Build-vs-Buy Equation
6. Shape Before Speed: The Four Core Principles of ForgeSDLC — doctrine-style hub tying the series to Forge’s four core principles

Citations across the series are summarized in the source reference list.

Four core principles (series)

Hub article introducing Forge’s four governing principles for AI-native delivery, with links to each deep dive:

• Shape Before Speed: The Four Core Principles of ForgeSDLC — full overview
• Shape before speed
• Flow over ceremony
• AI-first, human-gated
• Make gains compound

Forge principles (full list) covers execution principles and lean tenets—alongside these four core principles, not instead of them.

Autonomy and bounded execution

Sizing unattended agent work with cumulative gates—companion to AI-first, human-gated:

1. Autonomy Is Not a Switch
2. How Much Should an Agent Change Without You?
3. Verification Gates Scale With Autonomy

Handbook depth: Autonomy levels · Platform autonomy hub · Bounded execution examples.

More

• Why this blog exists

Agentic SDLC standout series

Human-oriented standout ideas promoted from the Forge Agentic SDLC hydration pack:

• Governed Delivery Control Plane, Not Generic Agent Orchestration
• ForgeRun Spine as the System of Record
• Agent Workcells as Replaceable Bounded Runtimes
• Approval Is Not Chat Memory
• Evidence-First SDLC
• L0-L8 Autonomy Ladder With Gates
• Defined, Demonstrated, and Vision Honesty
• Versonas as Callable Discipline Lenses
• Blueprints as Executable Policy, Not Static Docs
• LCDL as a Governed Reasoning Layer
• Fleet as an Approved Execution Rail
• Forge Campaign for Governed Multi-Repo Automation
• Micro-Agents and Micro-Packs
• Resource Honesty and the Worker Ladder

Platform architecture notes: standout hub.