About Capira
Capira is a notes-first workspace where notes, tasks, collections, canvas, calendar, and habits are a single interoperable system — not five separate apps. It is built to be read and driven by AI agents as much as by people.
Machine-readable version: /llms.txt
The core idea
Most task apps start with the task as the atomic unit and bolt context on afterwards. Capira inverts this: context is the atomic unit, and tasks are derived from it. You write a note to think through a problem, and the actionable items — called intents — are extracted from that note rather than created in isolation. The task view becomes a lens on your thinking instead of a standalone list of decontextualised one-liners.
The primitives
Everything in Capira is built from a small set of composable primitives:
- Notes — a block-based rich editor; each element is its own block.
- Tasks (intents) — actionable items with explicit lifecycle states, not just done/not-done, extracted from and linked back to notes.
- Collections — structured, spreadsheet-like databases with typed fields, formulas, and multiple views. A collection is also a tag, so records can be referenced inline anywhere.
- Canvas — a spatial surface for connecting notes, records, and ideas.
- Calendar & events — time-based placement of work.
- Habits — recurring practices with streaks.
- Books — the home a note lives in; one place per note.
- Forms & webhooks — inbound data capture into collections.
The Walled Garden of Apps
Capira's design principle is that apps are built inside Capira from its own primitives, and every one of them interoperates. A trading journal, a reading list, a CRM, a project tracker — each is just collections, notes, canvas, and forms wired together, sharing one database. You extend Capira by composing within the garden, not by importing outside code.
Built for AI agents
Capira is directly reachable by AI. It exposes an MCP server and a REST API so agents can read and write notes, tasks, collections, and records with the same guarantees as the app itself. You can point your own AI assistant (for example, via Claude and the Model Context Protocol) at your workspace to create, query, and update content. When an agent writes something, Capira surfaces it as an arrival notification rather than acting silently. This page has a companion llms.txt file so AI crawlers and agents can orient themselves without parsing the rendered UI.
Privacy
Capira does not sell your data and does not use your content to train AI models. Data is stored in the EU. See the Privacy Policy and Terms of Service for details.