Features

Neuronz.ai is a managed memory service for AI coding agents. It remembers what your agent learns, per project, and hands it back in later sessions. This page lists every feature, grouped by what it does.

Each page says what the feature does, how it works and how to use it.

New here? Start with Core concepts, or Connect your agent to get running.

The memory loop

What happens automatically in every session: relevant context goes in at the start and on each prompt, and what was learned is saved at the end.

• Persistent memory — short, single-claim facts, kept per project and cleaned of duplicates as they grow.
• Auto-recall — each prompt is searched against your stored facts, past work and docs, and the agent is told how much matches and how to read it.
• Session-start briefing — at the start of each session the agent gets your profile and pointers to your stored facts and rules.
• In-session self-consolidation — during the session, the agent saves what it learned itself, on your own subscription.
• Detached post-session capture — after you close a session, your agent saves anything still unsaved, in the background (Claude Code).
• Rules — rules you approve and the agent must follow. The agent is told which rules apply and fetches their text before it acts.
• Personas & the social map — what the agent knows about you and the people you work with.
• Voice — stores how your writing should read in different situations and applies it to drafts.

Knowledge and work

Records you write and read yourself, which your agent can also use.

• Knowledge docs — runbooks, decisions and reference material, split into sections so recall finds the exact passage.
• Topics & topic mode — ongoing issues and units of work, with priority, status and a kanban board.
• The knowledge graph — typed links between records, which recall follows to pull in connected context.

Agent enablement

What your agent gets in every session besides memory.

• Skills & assets — store skills, commands and subagents once; they appear in every session that should have them.

Platform

• Profiles — which folders share which memory, so projects never mix.
• Filling a profile — seed a new profile from your repositories, GitHub, Notion and docs, checked by a quiz.
• The plugin & automatic loop — the Claude Code and Oh-My-Pi plugins that run all of the above, with no per-repo setup.
• Plugin commands & skills — the commands you type, such as /neuronzai:capture, and the skills your agent uses on its own.
• Agent integration (MCP) — the ways an agent connects, and the MCP tools it calls.
• The dashboard — the web app where you read, search and edit everything.
• Usage analytics — how often your sessions use memory and your stored skills.

Your data

• Data sovereignty — who can see your data, and what it is used for.