A browser pool for agents that stop taking each other's tabs
Dan Mercede
Command Palette
Search for a command to run...
Archive
39 posts published so far
How My OpenAI Dot Starts Real Claude Code Sessions on My PC
Dan Mercede
How I Have Claude Code Control Grok Bot (There Is No API)
Dan Mercede
Your AI Pilot Did Not Fail. Nobody Owned the Workflow.
Dan Mercede
The Approve Button That Wasn't Attached to Anything
Dan Mercede
Why Most AI Governance Starts Too Late
Dan Mercede
Do My AI Coding Agents Actually Learn? I Built a Way to Measure It
Dan Mercede
The Fail-Closed Harness: Why a Safe Agent's Default Is Refusal
Dan Mercede
The Recon Refuted the Task: Building Infrastructure With an Agent You Can Trust
Dan Mercede
Quota Is Not a Parse Error
Dan Mercede
Refute Your Own Claims
Dan Mercede
Schema Is a Second Surface
Dan Mercede
The Router Is a Trap: Running Other Models Under Claude Code, Off Their Own Subscriptions
Dan Mercede
Certify vs Reproduce
Dan Mercede
Vet the AI PR Fleet
Dan Mercede
Giving Your Coding Agent Web Access Without Wrecking Its Context Window
Dan Mercede
Recon vs Prior Decisions
Dan Mercede
Cut Agent Web-Read Tokens
Dan Mercede
PR-Merge Train Gotchas
Dan Mercede
The Reviewer Blessed the Bug: Why Agent-Written Code Needs Layered Review
Dan Mercede
Benchmarks Evict Hot Models
Dan Mercede
Building a Governed, Double-Send-Safe Delivery Pipeline for Agent Outputs
Dan Mercede
Don't Index Your System of Record
Dan Mercede
The Limit Kills the Report, Not the Work
Dan Mercede
Silence Is Not Absence
Dan Mercede