---
id: "entity-stripe-minions"
type: "entity"
entityType: "tool"
canonicalName: "Stripe Minions"
aliases: ["minions", "Stripe minions"]
source_timestamps: ["§ Three Necessities", "¶15"]
tags: ["ai-agents", "software-development"]
related: ["concept-continuous-sensing", "action-analyze-task-level"]
sources: ["tail1"]
sourceVaultSlug: "hbr-seg-tail1"
originDay: 1
articleStem: "hbr-tail-112-continually-assessing-performance"
sourceUrl: "https://hbr.org/2026/06/the-pros-and-cons-of-continually-assessing-performance"
sourceTitle: "The Pros and Cons of Continually Assessing Performance"
---
# Stripe 'Minions'

**Entity type:** tool (internal AI coding agents at Stripe) · **Role in source:** evidence for Necessity #2 (task-level sensing).

Internal AI coding agents used at Stripe. In **February 2026**, Stripe reported that these 'minions' independently wrote blocks of code resulting in **over 1,300 submissions merged into production each week** — fully AI-written, human-reviewed. This gives Stripe granular data on the shifting division of labor: exactly *which tasks AI is absorbing* and *which engineers excel at supervising AI output*.

The minions are the cleanest illustration of [[concept-continuous-sensing]] producing a task-level division-of-labor trail, and they anchor the action [[action-analyze-task-level]]. Reading their output correctly requires the reframing in [[contrarian-productivity-vs-capability]] — value lies in *supervising* AI, not just accelerating output.