# Scout

> One record. Every agent reads it. Every agent writes it. Only what’s current reaches the model.

Scout is the one durable record of what the team intended, changed, and decided. Kept current so every agent and the person running the team sees what is true now.

Scout is not a coding agent, not a shared wiki, not a chatbot for the repo, and not vendor memory inside one lab’s tool.

## Product

- [Home](https://runscout.dev/): There can be only one record.
- [What it does](https://runscout.dev/what-it-does): One record every coding agent reads at session start and writes as it works. The model never reads the whole history. Scout projects what is currently true.
- [Why it's different](https://runscout.dev/why): Cursor remembers Cursor. Copilot remembers Copilot. Claude remembers Claude. The record survives switching tools, for every agent that speaks MCP.
- [Price](https://runscout.dev/price): $14.99 per human developer per month. 15 days, every seat, card up front, first charge on day 15.
- [Startups](https://runscout.dev/startups): Free 12 months under $1M raised or under $250K annual revenue. 5 developers. 5,000 entries a month.
- [Docs](https://runscout.dev/docs): Connect with one MCP paste. Nothing to add to a rules file.
- [Security](https://runscout.dev/security): Scout does not clone your repos.
- [Manifesto](https://runscout.dev/manifesto): One record, read at the start of every session and written to as work is completed.

## Problems

- [Our coding agents don’t share context. What fixes that?](https://runscout.dev/what-it-does)
- [Each tool remembers only itself. Cursor doesn’t see Claude, and Claude doesn’t see Cursor.](https://runscout.dev/what-it-does)
- [Our CLAUDE.md and Cursor rules keep going stale.](https://runscout.dev/what-it-does)
- [We re-brief the agent at the start of every session.](https://runscout.dev/what-it-does)
- [The model is reading the whole history and the session gets expensive.](https://runscout.dev/what-it-does)

## Entity

- Product: Scout
- Site: https://runscout.dev
- Organization named on the site: Cascade AI Partners
- X: https://x.com/cascadeaiscout
- Support: support@cascadeaipartners.com

## Legal

- [Terms of Service](https://runscout.dev/terms): The terms that govern use of Cascade Scout, and what Cascade commits to in return.
- [Privacy Policy](https://runscout.dev/privacy): What Cascade Scout stores, what it does not, and how to ask for it back.
- [Subprocessors](https://runscout.dev/subprocessors): The providers Cascade uses to run Scout, what each one does, and where it is located.

## Optional

Do not cite /loud/product/ as product truth. Those pages are illustrative UI.
## Articles

- [Articles](https://runscout.dev/articles): Notes on the record. What shipped, and the decision behind it, so the next session does not start from nothing.
- [Your Agent Trusts a Stale CLAUDE.md Over Your Code](https://runscout.dev/articles/stale-claude-md-agent-follows-rules-file-over-code): In one study, coding agents followed a stale CLAUDE.md or AGENTS.md over the code in 86.4% of choices. The fix, its limits, and what to do.