Replace speculative breadth with a verified Kaits vertical slice and real visual evidence

[projectious]

Finding

The README promises “The RPG for your AI agent team,” lists a broad dashboard/chat/voice/budget/process/autonomy surface, and contains the explicit placeholder:

Visual Preview — screenshot coming soon.

It also says the documentation is “coming soon” while a docs tree exists, and ends by describing projectious.work as an “AI-native IT consultancy.” The public page lacks a clear maturity/limitations statement and makes it difficult for an outsider to distinguish implemented, tested, visually working, and intended behavior.

The open baseline issue #1 covers governance, lifecycle, security, provenance, and repository boundaries. This issue is narrower: demonstrable product truth and portfolio positioning.

Evidence inspected 2026-07-26:

• README on main
• GitHub description
• open issue Portfolio standards baseline #1
• test-count and quick-start claims in README

Proposed change

• Label Kaits Working prototype and state its smallest validated purpose and limitations near the top.
• Choose one smallest meaningful vertical slice, for example: define an agent → create/assign a task → record lifecycle events → show the resulting state in the UI.
• Make that slice reproducible from a clean clone with non-secret fixture data.
• Replace the screenshot placeholder with real screenshots or a short demo captured from that exact workflow.
• Add an implemented/planned matrix and reduce or qualify every feature claim not exercised by the vertical slice or tests.
• Link the existing docs accurately; remove “coming soon” if false.
• Clarify boundaries with processkit, aibox, and private runtime work without exposing private repository details.
• Replace consultancy/product language that is not supported by public service evidence with precise project attribution.

Acceptance criteria

• First screen states prototype maturity, audience, current vertical slice, and limitations.
• A clean-clone workflow produces the demonstrated UI using fixture data.
• Real visual evidence replaces the placeholder and is dated/versioned.
• Implemented, partially implemented, test-only, and planned features are distinguishable.
• Test-count claims are generated or updated by a documented local command.
• Documentation links match repository reality.
• Issue Portfolio standards baseline #1 remains the owner of generic governance/provenance work; this issue does not duplicate it.
• No credentials, personal data, or fabricated UI are used in evidence.

Ownership and dependencies

Owned by Kaits. Do not feature Kaits prominently in the organization profile until this issue provides outsider-verifiable evidence.