Gradient Push

Gradient Push

AI and automation for people who build things. Weekly.

Agent rollback drill

Jun 25, 2026

•

1 min read

Agent rollback drill

A 10-minute drill for checking whether an agent can be paused, rolled back, and explained before it gets more access.

Gradient Push
Gradient Push
Score one agent permission before you allow it

Jun 20, 2026

•

1 min read

Score one agent permission before you allow it

Five checks for deciding whether an agent action can run, needs approval, or should stay blocked.

Gradient Push
Gradient Push
Before your agent gets more access

Jun 14, 2026

•

1 min read

Before your agent gets more access

A short checklist to copy before an agent gets permission to change things outside its sandbox.

Gradient Push
Gradient Push
Before your agent gets more access

May 30, 2026

•

2 min read

Before your agent gets more access

If an agent can send, change, refund, merge, or deploy, check the tool boundary before you expand access.

Gradient Push
Gradient Push
MCP Servers Need Trust Boundaries, Not More Connectors

May 24, 2026

•

6 min read

MCP Servers Need Trust Boundaries, Not More Connectors

MCP makes agent integrations easier to compose, but every new server also adds a trust boundary. Production teams need ownership, scopes, audit evidence, failure modes, and revocation paths before connector sprawl becomes agent infrastructure risk.

Gradient Push
Gradient Push
Self-Healing Tests Need an Operator, Not a Vibe

May 19, 2026

•

6 min read

Self-Healing Tests Need an Operator, Not a Vibe

AI-assisted test repair can reduce maintenance toil, but only if teams define what may heal automatically, what requires review, and what evidence proves the test is still protecting the behavior users depend on.

Gradient Push
Gradient Push
Load more