# Notiky

> The product engineering brain — persistent decision/hypothesis/insight graph for teams who build software with AI agents.

Notiky captures the *why* behind every product and engineering decision, then feeds that context into Cursor, Claude Code, Codex, Gemini CLI, and your team. It is a product engineering brain, not another task tracker.

## What Notiky is

- **Product engineering brain** — the decision, hypothesis, and insight graph that captures *why* every call was made, not just *what* was built.
- **Persistent context layer for AI agents** — structured product memory that coding agents consume via MCP, daemon, and exported artifacts (CLAUDE.md, skills, rules).
- **Human + agent coordination layer** — humans and coding agents work on the same decisions, hypotheses, and tasks with shared context.
- **Context delivery system** — every agent run can open with a context pack (CLAUDE.md, linked decisions, team skills) so agents arrive with product memory, not a blank prompt.

## Who it is for

- Founders and solo builders who use AI agents daily and keep re-explaining context
- Product managers and tech leads who lose the "why" behind decisions within weeks
- Small engineering teams (2–10 people) running multiple coding agents in parallel

## What it is NOT

- Not a task manager or issue tracker (Linear, Jira do that well — Notiky is the layer above)
- Not a generic chat tool (it's a structured brain with provenance, not a conversation log)
- Not a note-taking app (decisions have evidence, supersession trails, and links to code)

## Key Features

- **Copilot (Ask / Think / Plan / Do)** — operator sessions for product and engineering work; Think and Plan modes stress-test assumptions and produce structured outputs tied to the brain.
- **Decisions** — first-class objects with linked hypotheses, evidence, rejected alternatives, and supersession trails. Reopen any decision and the full context is there.
- **Tasks + todos** — work units derived from the brain; agents execute with workspace context via the local daemon.
- **Autonomous runs** — agent runs can ship with a context pack. Blockers surface early. Runs are replayable.
- **Artifacts** — project-specific coding infrastructure (CLAUDE.md, cursor rules, UI guidelines) generated from current decision state.

## Supported Coding Agents

Claude Code, Codex, Cursor, Gemini CLI, Copilot CLI, OpenClaw, OpenCode

## Privacy & Security

Code never passes through Notiky servers. Agents run on your machine via the local daemon. Notiky coordinates context and state — your code, decisions, and data stay in your environment.

## Pricing

See /pricing.md for machine-readable pricing data.
Current tiers: Free, Basic, Plus, Ultra.
Pricing page: https://notiky.com/#pricing

## Key URLs

- Homepage: https://notiky.com
- About: https://notiky.com/about
- Changelog: https://notiky.com/changelog
- Download (CLI + desktop app): https://notiky.com/download
- Pricing: https://notiky.com/#pricing
- Machine-readable pricing: https://notiky.com/pricing.md

## Social

- X (Twitter): https://x.com/notiky_com

## Contact

- Website: https://notiky.com