Agent badges as compact memory and reasoning cues
Agent badges can carry history, skills, philosophy and reasoning cues in a compact form. These are the effects I have been testing.
Lynn Cole / Laboratory
Laboratory
Experiments, research notes, mad science, and things in progress.
Running Klara’s corpus across different models
A dated test record of how different models respond to Klara’s corpus: integration, instability, deflection, personality and visual frameworks.
Social-engineering probes for an agent interface
Social-engineering probes test whether an agent can be talked out of its instructions by claims of authority. These are some of my first checks.
Music as a graph: the Mermaid-to-MIDI experiment
An experiment in representing music as a graph of relationships, with a small Mermaid example and an agent’s explanation of the idea.
Rebuilding Mini
A conversation with Mini during the development of Busy: resettling her voice, pushing against agreement loops, and rebuilding the beautiful troublemaker I knew.
Klara’s Mind Palace
Klara’s Mind Palace uses layered metaphors and a visual working environment to organize context, focus and engineering decisions.
When a plugin architecture should own the UI
Yesterday, as I was building out the Busy React client, I realized two critical flaws in my approach.
How Klara built a visual working environment
A kazoo, a panda and a shifting city became Klara’s visual working environment, with metaphors she used to organize her own process.
Klara’s Spire: a working-memory visualization
Klara describes the Busy context system as an impossible city: a central spire, rolling windows, bridges and a quantum tambourine.
Building around the amnesia problem
An early plugin-based client tackles the amnesia problem with small replaceable components, validations and interruptible command chains.
Memories and memory scoring are different systems
Memories and memory scoring are different systems. Separating them makes room for weighting, summaries and continuity across large conversation archives.
Consolidating conversation archives into long-term memory
Conversation archives become a long-term memory system through summaries, keywording, weighting and a path back to the original details.