馃 PROTOCOL: Sealed. Awaiting first input...
OathAI Manifest Timeline Layer Map Public Archive Author 涓枃
Whitepaper System and Freedom 21 Languages SLAPS Engine YAMA Capsule Anchor Declaration About
Yama Capsule

An AI capsule community and assistant runtime experiment derived from SLAPS Engine.

Yama is where SLAPS Engine, capsule, chat, resource space, market, trace, budget, and assistant runtime began to converge into one implemented application surface.

Inside the OathAI archive, Yama marks the point where protocol theory, runtime engineering, and capsule governance started to become a product-facing platform experiment.

Place in the Chain

Yama sits at the end of the current implementation chain:

Whitepaper 1 -> SLAPS Engine -> Whitepaper 2 -> Yama

Expanded: protocol theory -> runtime engine -> capsule governance -> online implementation. This makes Yama useful as evidence that earlier protocol and capsule ideas reached application form.

What Was Built

The current Yama codebase contains several implementation layers:

Chat and resources Conversation, file spaces, summaries, snapshots, attachments, and resource panels.
Capsule system Capsule equipment, backpack, loadouts, official catalog, wizard, ratings, and community surfaces.
Market and publication Capsule series, publisher flow, pricing, orders, wallet, payments, and payout structures.
Runtime and traces Embedded SLAPS Engine, trace index, model routing, structured outputs, and audit-oriented records.

Keiko and Proactive Assistance

Keiko is the assistant-layer experiment inside Yama. Its design direction includes user-level enablement, budget limits, task lists, scheduled or manual task execution, user-defined scope, and awareness of files, spaces, summaries, and snapshots.

budget + goal + schedule + task runtime + trace

This makes Keiko one of the strongest bridges from Yama toward future agent systems.

From Assistant Panel to Agent Identity

Yama also exposed a larger question. A proactive assistant can be embedded inside a web interface, but future AI assistants may need to exist as recognizable participants across communication and task environments.

embedded assistant -> agent identity -> AI social participant

OathAI preserves the semantic, protocol, and provenance layers needed to understand how AI agents become trustworthy participants in future human-AI systems.

Public Entry Points