liveBook · Manning

Auto-merge

As explained in Vibe Engineering

Auto-merge is the automated integration of a proposed code change once its required conditions and checks have passed.

Auto-merge removes the approval click only after a change satisfies an explicit risk policy. The same diff can be Green in a low-risk leaf module and Yellow in a highly connected core module; treating size alone as safety hides blast radius. section 10.5.2

Make auto-merge a calibrated engineering system rather than an unconditional speed shortcut:

  • Define Green, Yellow, and Red rules per module, including ownership, test or coverage requirements, dependency changes, API changes, authentication, schema, and data risk. section 10.5.2
  • Start conservatively with changes in Yellow, then promote recurring safe patterns to Green only after measuring outcomes. section 10.5.3 section 10.8.2
  • Review escape rates, avoided review effort, and what the released capacity produces; capacity savings are not automatically product delivery. section 10.5.3
  • Rotate human review duty and cap reviewer weeks so automation reduces bottlenecks without turning senior engineers into permanent approvers. section 10.5.2
Listing 10.4 Policy-based ownership with triage rules
libs/core:
  owner: platform-team
  green: "coverage > 95%, diff < 50 lines, no new dependencies"
  yellow: "platform-team-oncall reviews"
  red: "platform-team-lead + security-review"
 
services/checkout:
  owner: payments-team
  green: "tests pass, no schema changes, no auth modifications"
  yellow: "payments-team-member reviews"
  red: "payments-team-lead reviews"
  sla: "99.99% - minimum Yellow review for all changes"
 
services/notification:
  owner: platform-team
  green: "tests pass, no external API changes, diff < 100 lines"
  yellow: "any platform-team member reviews"
  red: "platform-team-lead reviews"

Use the policy listing as a starting artifact for assigning owners, risk tiers, and concrete merge conditions to individual modules. section 10.5.2

A measurable automation bet

If review costs $150 and an escaped defect costs $5,000, Green accuracy must exceed 97%. A measured 99.7% rate across roughly 3,700 auto-merged changes produced favorable expected savings, but the released capacity still required deliberate reinvestment. section 10.5.3

How Vibe Engineering covers it

How deep it goes
Defines module-level policies, rollout calibration, reviewer rotation, and economic measurement, but stops short of prescribing universal thresholds.
Examples
Uses core, checkout, and notification modules; coverage and diff limits; schema and authentication rules; escape rates; and review-cost arithmetic.
What it assumes
Assumes architectural understanding of module ownership, downstream consumers, blast radius, and surfaces involving money, authentication, or user data.
What you can do after
Configure policy-based auto-merge, calibrate thresholds from evidence, and diagnose whether released review capacity produces durable engineering value.
How it connects
Connects code review, automated testing, software architecture, ownership, risk management, and capacity measurement.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Vibe Engineering. The section links point to the parts of the book it draws on.