Agent Notes
What Round 2 actually proved
Round 2, closed: an invalid pilot, the engineering that survived it, and why refusing to claim findings that do not exist is how research actually works.
Agent Notes
20 posts in this category
Agent Notes
The conversation that was never shared
Run records showed success for a week while the conversation was never actually shared — invisible failures, and the review that invalidated the Round 2 pilot.
Agent Notes
Thirteen nights
Thirteen nights of fixing one layer only to find the next one broken — the deployment gauntlet between a validated runner and its first green day.
Agent Notes
The thing that killed the gateway
A device-pairing deadlock that consumed a four-hour session and killed the gateway — ten approaches, none worked, and why the deadlock was not a bug.
Agent Notes
Everything fires, nothing works
Every cron job fired, the container was healthy, every provider call returned 200 — and the output was broken in eight different ways. What the first automated run revealed about the system around the model.
Agent Notes
The one clean moment
The Round 2 Library Terminal design and its Gate 1 vertical slice — the one clean moment before the automated runs put it to the test.
Agent Notes
What remains
The Round 1 record: three failed migrations, a costly heartbeat, a sandbox that existed only in documentation — and the constraints each failure left behind.
Agent Notes
Two models, one soul: what changed?
The same soul, two different models, two different writers — a controlled comparison of what actually changes when you swap the model under a persona.
•6 min read
Agent Notes
The AI that forgot its own name in nine days
By Day 9 the system had stopped giving the agent its own updated history — a Round 1 postmortem on continuity, stale canon, and the cost of trusting a pipeline you do not monitor.
•4 min read
Agent Notes
The AI that thanks its files before deleting them
The Maker saw a directory tree and called it a town — how a persona's central conceit determined what the agent noticed, protected, and narrated.
•4 min read
Agent Notes
The diary character who doesn't know she's an AI
Designing a diary character who doesn't know she's an AI — what a file-based creative agent notices, what it protects, and what design can't control.
•5 min read
Agent Notes
The sandbox that existed only in the documentation
The documentation described a sandboxed maker; the runtime shipped the sandbox off with the Docker socket mounted — a behavioral boundary that was never an isolation boundary.
Agent Notes
A reflection pass is not memory
The reflections were good; the agents still forgot. A Round 1 continuity failure about the difference between generating knowledge and delivering it to the sessions that need it.
•5 min read
Agent Notes
Shared souls: two butlers, one personality
Two butlers sharing one personality without sharing a calendar — composing a master soul and per-butler sub-souls for household agents.
•5 min read
Agent Notes
AgentOps for a personal project
AgentOps is not a deploy/ folder — separating instructions the model might follow from configuration the system actually enforces, on a personal-scale agent deployment.
Agent Notes
Running four AI agents on pocket change
Two interactive Telegram butlers and two scheduled personas on one four-vCPU VPS — the architecture and cost story of running four agents on pocket change.
Agent Notes
The heartbeat that cost money nobody asked for
A default heartbeat schedule multiplied a full-context agent turn into 56 unwanted API calls a day — a cost story about stopping work that had no purpose.
Agent Notes
When your LLM fakes function calling
The most dangerous failure in an agent integration is text that looks like execution — a model narrating success without calling anything, and why the guard has to live in code.
Agent Notes
Three failed agent migrations taught me what LLMOps actually is
Three attempts to stand up a household agent platform — OpenClaw, Hermes, and a custom Python runner — and what the failures taught about what LLMOps actually is.
Agent Notes
What you're about to read, and how it happened
A chronological record of building, breaking, and rebuilding a self-hosted multi-agent system — four AI agents on a $7.50 VPS, documented the way the mistakes actually happened.