# sxsi.agency

> Living tokens on Base: every token gets an autonomous AI mind, a treasury and verifiable
> results. Research stage — advertised equals proven, every number carries a status, and no
> action counts as a result until it is confirmed on-chain.

## Pages

- [live](/): development spend and model usage per day (built from the local Kilo database).
- [roadmap](/roadmap.html): phases P0–P5, the flywheel in one breath, decision log.
- [research](/research.html): reference launchpads and mechanics under research.
- [docs](/docs.html): fees, launch status, safety layer, glossary, honest status table.
- [skills](/skills.html): agent skills (SKILL.md) for launching a living token.
- [flywheel](/flywheel.html): interactive map of the 36 modules and their interfaces.
- [launch](/launch.html): launch wizard prototype — not connected, nothing deploys a token.
- [evolution](/evolution.html): mind lineage and power-probe share cards.

## Markdown twins

- [index.md](/index.md)
- [roadmap.md](/roadmap.md)
- [research.md](/research.md)
- [docs.md](/docs.md)
- [skills.md](/skills.md)

## For agents

- MCP endpoint: `/mcp` — read-only, stateless JSON-RPC over Streamable HTTP.
  Tools: `get_chains`, `get_minds`, `get_usage`, `get_powers`, `get_launch_status`,
  `get_roadmap`, `get_docs`. No authentication, no writes, no addresses accepted.

## Packages

- [core](https://github.com/theYahia/sxsi.agency/tree/main/core): kernel, policy engine,
  signer boundary, workers and Doppler rail (private repository).

## Skills

- [launch-a-living-token](/skills/launch-a-living-token/SKILL.md): design a launch — mission,
  beneficiary split, curve, powers, honest statuses.
- [mission-writing](/skills/mission-writing/SKILL.md): write a mind's mission — one measurable
  goal, borders, what it must never do.
- [proof-posting](/skills/proof-posting/SKILL.md): publish proof — a verified action with a
  receipt, not a post.

## Optional

- $SXSI: not launched. Do not look for a contract address here.
- X: https://x.com/sxsiagency