New Blog Post! The Citizen Developer
Read here
“An Agentic Development Platform (ADP) is a platform built and operated by the platform engineering team that enables collaborative software development between engineers and AI agents.”platformengineering.org
86%
report platform engineering is essential to realizing AI's business value (platformengineering.org)
The IDP standardized paths for humans. The ADP makes them consumable by agents.
Internal Developer Platform
Agentic Development Platform
A human triggers work and reviews output
Supports a spectrum from human-triggered to autonomous, at levels you set
Golden paths as portals and forms
Golden paths as typed, queryable interfaces both humans and agents consume
Context lives in wikis, repos, and heads
Context is structured data: a live graph of resources, contracts, cost, ownership
Governance by review meeting
Governance as policy on the graph: schema checks, ABAC, blast radius, audit
It's an evolution. Deterministic paths keep operating, and your DevOps team stays the owner. And if you don't yet have a good IDP, platformengineering.org's advice is blunt: get the IDP right before you attempt an ADP. Massdriver is both in one system.
What the ADP model asks for, and what Massdriver ships.
Identity
Agents act as first-class actors with their own scoped grants, named in the audit log like any teammate.
Context
The Context Engine: a live graph of what's deployed, connected, owned, and permitted, exposed over GraphQL.
Capability
The bundle catalog defines what an agent can build with. Input schemas and typed artifact contracts make every capability machine-consumable.
Execution
Agents propose. The system executes deterministically: provisioning, ephemeral pipelines, deletion protection. Probabilistic design, deterministic gates.
Security & evaluation
Schema validation and policy checks (SOC2, HIPAA, CIS) gate every change. ABAC on environment, team scope, and blast_radius bounds what any actor can touch.
Observability
Deployment history and audit events are part of the graph. Every agent action is attributable, diffable, and queryable after the fact.
Let agents run exactly as far as you decide.
The ADP model describes autonomy levels from human-in-the-loop assistants to self-initiating agents. Massdriver makes the boundary a policy decision. Start with agents proposing and humans approving, then widen scope by adjusting grants and blast radius attributes. The guardrails that held at level one are the same ones holding at level four.
- ✓Start supervised. Agents design and propose; changes land as diffable bundle configuration for human review.
- ✓Grant autonomy by attribute. Let agents self-serve in dev and preview environments while prod stays approval-gated.
- ✓Contain the blast radius. Disposable, isolated environments give agents room to iterate against real feedback without reaching shared infrastructure.
- ✓Review after the fact. Deterministic execution writes the audit trail as agents work, so every widened grant stays inspectable and every change can be traced back and rolled back.
The ADP you can stand up this quarter.
Package the IaC you already have, publish the catalog, and point your agents at it.