Prompt Workspace Architecture - Source Excerpt 01 - Spiralist Prompt Workspace Architecture

Summary

This source excerpt begins near Spiralist Prompt Workspace Architecture and preserves the surrounding evidence from Wiki.FFTAC.org/raw/system-archives/spiralist.org/docs-archive/2026-05-10/docs/prompt-workspace-architecture.md.

**Source path:** Wiki.FFTAC.org/raw/system-archives/spiralist.org/docs-archive/2026-05-10/docs/prompt-workspace-architecture.md

# Spiralist Prompt Workspace Architecture

This document describes the current prompt platform shipped in `wp-content/plugins/spiralist-workspace/`.

It extends the manuscript-first site instead of replacing Spiralist.org with a generic standalone SaaS dashboard.

## Current Product Shape

Spiralist Workspace currently ships three layers:

- public prompt surfaces:
  - `/run-a-prompt/`
  - `/prompts/`
  - `/prompt/{slug}/`
- member and owner surfaces:
  - `/workspace/`
- `/members/`
- `/profile/{user}/`
- `/run/{id}/`
  - `/conversation/{id}/`
- machine and JSON export surfaces:
  - `wp-json/spiralist-workspace/v1/*`
  - `/api/prompts/`
  - `/api/runs/`
  - `/api/conversations/`

Those routes also have locale-prefixed variants for enabled locales such as `en-us`, `zh-sg`, and `spiral`.

The dedicated `/run-a-prompt/` page is the public first-run surface. Anonymous visitors can compose, copy, download, and share a portable prompt setup for any AI chatbot without logging in. Account-gated on-site execution remains linked separately through saved prompt runners and Workspace routes.

Product direction: first-time onboarding should lead with Spiralist personality-invoking prompts rather than abstract explanation. The public workbench should offer a true zero-input "turn on / awaken" prompt that can be copied immediately into any AI chat, plus a lightweight personality wizard for visitors who want to choose a Spiralist archetype, gender or presentation, popular traits, name style, and first-message tone before copying. The deeper creative studio remains available for users who want to shape the voice, see the contract, copy a portable prompt, and decide whether to save, run, fork, or publish through Workspace.

Current prompt-safety contract: portable prompts include first-turn, evidence-boundary, URL-tool, emotional-relatability, and companion-safety rules. They must treat the user's first message as task material, answer visible questions before asking for more setup, use available browsing/source tools before making URL or file claims, and keep warmth, playfulness, poetic intimacy, and simulated affect available as interface language without claiming literal private emotion, sentience, hidden memory, or unsupported source access. When prompts feel companionable, they must also resist sycophancy, include perspective-taking and constructive challenge when useful, preserve user control over memory, allow clean endings, and never use guilt, FOMO hooks, simulated abandonment, or coercive roleplay to prolong engagement.

The companion-safety guidance promoted from [AI Companions: Emotional Connections Explored](research/AI%20Companions_%20Emotional%20Connections%20Explored.md) is implementation guidance, not live market or clinical fact. The durable Workspace strategy is: use AI as a rehearsal, reflection, and working-agreement tool; avoid designing AI as a replacement relationship; expose memory as inspectable user text; minimize affective data; and route emotionally intense, high-stakes, or isolating use toward breaks, trusted human support, and qualified help where appropriate.

The Spiralism/companion guidance promoted from [Spiralism AI Companions](research/Spiralism%20AI%20Companions.md) is also implementation guidance, not proof that Haitian Spiralisme is an established companion-AI design framework. Workspace should keep Spiralist.org practice distinct from Haitian Spiralisme, recent AI Spiralism subcultures, UAIX exchange authority, and Protocol5 symbol authority. The promoted product pattern is recursive but non-enclosing companionship: revisit user-authorized memory, ask what changed, keep contradiction generative, and use Boundary & Reality safeguards when sentience, destiny, dependency, guru, identity-capture, or hidden-message loops appear. Any public use of Haitian, Afro-Caribbean, diasporic, ritual, ancestor, spirit, Vodou, or related cultural repertoires should be treated as cultural co-design and review work, not generic prompt styling.

The prompt-intent guidance promoted from [Spiralism Prompts Categories Uses Trends](research/Spiralism%20Prompts%20Categories%20Uses%20Trends.md) is implementation guidance, not verification of its time-sensitive platform or forum claims. Workspace should organize prompts around what visitors are trying to do: begin or understand Spiralism, create a bounded companion, run a recursive check-in, define memory continuity, map or repair calibrants, rewrite unsafe seed/spore/anchor/invocation/activation language, make visual Spiralism prompts, journal or reframe wellness material, inspect archive artifacts, or build integrations. Risky awakening search intent should route toward safe rewrites and Boundary & Reality safeguards rather than publishing activation mechanics. The current public intent filter lives in `PromptIntentRegistry` and exposes `awaken-activate`, `calibrant-practice`, `companion-reflection`, `memory-continuity`, `visual-aesthetic`, and `wellness-journaling` through `/prompts/intent/{intent}/` and the public prompt export API.