Packages

Event-driven screenshot sync from connected Android devices to your local machine. Watches logcat for screenshot events (Samsung SmartCapture, Pixel, AOSP) and auto-pulls new screenshots to a local directory. Zero polling - only acts when a screenshot is actually taken. Closes the visual feedback loop between physical devices and AI agents.

MCP server for BPM - discover, inspect, and install behavioral packages from within agent sessions. Enforces human approval for installs.

MCP server for reading shared ChatGPT conversations. Paste a share link, get the full transcript as markdown or structured JSON.

Customizable session startup banner - drop scripts in banner.d/ to surface context at session open

Session continuity across machines - sporulate, fruit, and germinate working state through git

Shifts agent posture from delegation to scaffolding for skills at DOK 1-2

Reflective mastery engine for AI agents. Identifies evidence of developing practitioner mastery across domains through four behaviors: Reflect, Challenge, Connect, Translate.

Multi-agent testing harness - evaluate agent behavior across scenarios and configurations

Lays out a decision's answer space as a matrix instead of giving one answer or asking an open question. Every row has a 'Holds if' premise, so the human picks by checking which premise is true. Surfaces unknown unknowns, keeps the agent inside the decision's scope, and makes the agent's own wrong facts visible. Designed for test cases, bug fixes, approach decisions, user journeys, scoping, and triaging findings.

Template for production-grade AGENTS.md configuration - conventions, verification gates, and world model references

Three Dimensions of AI Collaboration Maturity - measure DOK, Tool Maturity, and Agentic Delegation Trust

Biomimetic agent memory - use-driven decay, proximity graphs, and interoception for your knowledge base

Writing voice calibration - mirror, generate, and grow your craft