Agent Memory & Skills

How a hybrid Rust and Solana architecture gives AI agents mathematically verifiable memory without sacrificing speed.

A tiny metadata layer that turns a content site into an agent-friendly knowledge base, with `llms.txt` for language models and CSV for pipelines.

vtcode-acp-client is a tiny thread-safe seam for sharing a single Agent Client Protocol connection across a Rust agent stack, without passing state through every layer.

A browser-extension-first monorepo that scrapes pages, builds embeddings in the background, and uses PostgreSQL plus pgvector to turn saved links into retrievable knowledge.

A first look at how Rhino Health is positioning skills as the layer that lets general AI agents plan, execute, and stay inside the rules of federated healthcare workflows.

A deep dive into the warm pool, session durability, Markdown memory, and approval gates that turn stateless model calls into inspectable, stateful agent sessions.

A Bash-based runtime uses Lark Base, Messenger, and Approvals to give AI agents memory, workflow, and guardrails without building a separate platform.

Daksh Jaitly’s Python project turns Slack, Discord, and Telegram into interchangeable entry points for one persistent assistant with local memory and state.

A Next.js showcase where <code>.agent/skills/taste-skill/SKILL.md</code> acts like a design system for the coder, not just the interface.

Inside Tank’s security-first skill registry, where agent instructions become versioned, permissioned, and audit-ready instead of living as brittle prompt advice.

Graphite Memory blends graph and vector retrieval with decay, suppression, and scoped sessions so an agent remembers what matters, not everything it has seen.

A plain Git repo can do something most note apps cannot: make your thinking portable, inspectable, and recoverable.

How a pipeline of 25 modular Markdown files gives local-first AI a continuous heartbeat, durable memory, and a strict metadata firewall.

How a local-first middleware uses strict human breakpoints to turn unpredictable LLM execution into a deterministic state machine.

AI agents are excellent at extracting business logic from legacy code, and terrible at getting it completely right. Here is how a hybrid Markdown-and-SQLite architecture forces human consensus before AI guesses become organizational truth.