Change-Intent Governance: The Missing Layer

说明与用途

实施变更意图治理层提供不可变、可审计的变更影响和批准记录，使即使在 AI 代理大量生成代码的情况下也能进行可靠的事后审计和合规性检查。

问题

Organizations lack a layer that records which changes affect which projects, who approved them, and the evidence, especially as AI agents generate most code.

结果

Implementing a change‑intent governance layer provides immutable, auditable records of change impact and approvals, enabling reliable post‑mortems and compliance even as AI agents generate code at scale.

工作流

• 1. Capture a portfolio snapshot (projects, dependencies, source‑of‑truth) as immutable evidence.
• 2. Compute impact of a proposed change from the snapshot’s dependency graph.
• 3. Record the change request, approval/rejection, and any remediation plan as an append‑only entry with actor attribution.
• 4. Execute the change via runner sessions, logging each action and its evidence.
• 5. Store all records in a read‑only ledger (SQLite) that cannot be altered after creation.

证据

Both layers are real and both are worth having. But there is a concrete question neither of them can answer: this change that is about to land — which projects does it affect, who approved it, and where is the evidence?

来源

I maintain an open‑source reference implementation of this layer: aine-control-plane… It is a dependency‑free Python core… append‑only SQLite evidence and audit storage…

来源

When a change lands across more than one repository, is its blast radius recorded, immutably, as a claim someone approved — or only discoverable, after the fact, by an engineer reading diffs?

来源