Personal AI Orchestrator

Goals → tasks → agents → results

Exocortex is your external brain. It understands intent, pulls the right context from your 2nd Brain, assembles the right AI agents, and ships outcomes — with logging, feedback loops, and smart time planning.

Why Exocortex

We shorten the path from idea to shipped outcome — tuned to your context, time, and goals.

Intent → Plan

Understands your goal, drafts a plan, and breaks it into actionable tasks.

2nd Brain integration

Works with your memory — notes, files, knowledge graph — no copy-paste.

Agent orchestration

Selects and coordinates specialized agents to execute each step.

Auto-scheduling

Respects your calendar, proposes slots, and balances priorities.

Learning loop

Captures feedback and improves the next iteration automatically.

Privacy by design

Local-first mindset with explicit permissions. You choose what to connect.

How it works

Minimal magic. Maximum transparency.

Tip

Exocortex removes friction between thinking and doing: the system picks the right agents and wires them to your context — you keep control over each step.

Core modules

The foundation for the MVP and the growing ecosystem.

2nd Brain (Knowledge Graph)

A living graph of ideas, tasks, links, and sources.

Composer-0

Planner & orchestrator: builds pipelines from agents and tools.

Context retriever

Grabs the right context from notes, calendar, files, and the web.

Task processor

Executes steps, manages statuses, logs, and artifacts.

FreeMinder

Goals & planning: prioritization, time slots, and focus mode.

Agent marketplace (next)

Plug external micro-capabilities with ratings and incentives.

Use cases

Designed for tech-savvy professionals.

Research → Plan → Deliver

Gather sources, synthesize, plan, and ship the final artifacts.

Content pipeline

Brief → drafts → editorial → publishing schedule → outcomes.

Product ops

Backlog → decomposition → agent orchestration → calendar integration.

Deep dive

More context for who it helps and what you actually get.

Who it's for

  • Builders who ship (PMs, ML/AI engineers, indie hackers).
  • People juggling research, writing, planning, and delivery.
  • Teams that want reusable, transparent agent pipelines.

What you get

  • Faster path from intent → finished artifact.
  • Memory graph (2nd Brain) so you don't repeat yourself.
  • Calendar-aware "next best action" and clean logs/artifacts.

Example pipeline: Research → Outline → Draft → Refine

  1. Intent: "Write a 1-page brief on X." Definition of done set.
  2. Context: pull prior notes/sources from 2nd Brain + fresh web cites.
  3. Compose: Composer-0 chooses agents (summarize, outline, draft, polish).
  4. Execute: steps run with artifacts saved (sources, outline, drafts).
  5. Schedule: slot offered for review/edit based on your calendar.
  6. Review: quick feedback; final PDF/markdown shared with provenance.

Module details

Precise roles and I/O so the system stays understandable.

Preminer (intake & triage)

Single inbox for thoughts/notes/commands. Classifies (task, note, idea) and routes.

  • Inputs: free text, voice, labels
  • Outputs: tasks to plan, notes to 2nd Brain, pipeline requests to Composer-0
2nd Brain (knowledge graph)

Living graph for People/Projects/Tasks/Notes/Sources with timeline + semantic search.

  • Ops: add/merge entities, link relations, attach artifacts
  • Retrieval: hybrid graph walk + dense search with citations
Context Retriever

Fetches precise, permissioned context from 2nd Brain + calendar/files/web.

  • Typed queries; freshness/quality scoring
  • Returns snippets + citations; scope-aware
Composer-0 (planner & orchestrator)

Turns intent into an agent pipeline with steps, deps, and acceptance criteria.

  • Templates for research/content/code
  • Chooses agents/tools; keeps it observable
Task Processor

Executes steps, retries safely, logs everything, collects artifacts.

  • Idempotent runner; event log
  • Artifacts vault (files/links) with provenance
Planning System

Suggests time slots respecting work hours, soft blocks, and energy profile.

  • Models: WorkHours, TimeBlock, EnergyProfileEntry
  • API: suggest_slots() → ranked windows
Agent Layer / Marketplace

Pluggable micro-agents with typed I/O, ratings, budgets, and permissions.

  • Clear scopes & resource caps
  • Future: community agents with governance

FAQ & privacy

How is this different from a task manager?

It composes agent pipelines from your intent and context, executes steps, and logs artifacts. Tasks are the by-product, not the product.

Where does my data live?

You explicitly connect sources. 2nd Brain keeps a permissioned memory graph with citations and an audit trail.

Can I plug custom agents?

Yes—an Agent API is planned for beta. Agents have typed I/O, scopes, and budgets.

Can it schedule for me?

The Planning System proposes slots that respect your work hours, soft blocks, and energy patterns.

What do I get in the beta?

Saved pipelines (research/content), artifact vault, calendar-aware execution, and early access to the Agent API.

Pricing?

Early beta is free for waitlist users. Paid tiers will reflect agent usage and collaboration features.

Roadmap

  1. MVP

    Landing + waitlist, basic task orchestration, 2nd Brain integration, calendar slots, artifact logging.

  2. Beta

    Agent marketplace, personal profiles, improved retrieval, pipeline templates, private instances.

  3. Vision+

    Thinking OS: personal task bus, multi-agent scenarios, live knowledge graph, self-learning.

Join the waitlist

Become an early user/builder — get invites, updates, and access to private builds.

Or email us at info@exocore.cx