# Workspace Through WaterStorage

This is a first-pass hydrological reading of the workspace at `/home/workspace`.

## Ocean

The raw signal layer: logs, transcripts, source files, open-ended drafts, chat remnants, public web saves, and all incoming material that has not yet been shaped into a stable artifact.

Examples:
- `AnthroCybernetics/sources/water-storage.md`
- `AGENTS.md`
- `conversation-summaries/`
- raw project notes and transcripts

## Evaporation

Abstraction and compression.

This happens when long documents become framework summaries, when a sprawling project becomes a source-of-truth document, or when a working session gets distilled into an index, README, or seed.

Examples:
- a course manual becoming a compact source map
- a chat session becoming a summary or handoff
- a project folder being reduced to a canonical README

## Clouds

Shared context distributed across the workspace.

Examples:
- `AGENTS.md` files that route understanding across folders
- `SOUL.md` when present
- recurring patterns such as Cascade, Wholewheat, Medulla, WaterStorage, Honeycomb, and the Anthrocybernetics course
- cross-project terminology that multiple artifacts share

## Rain

Triggered landing of context into a concrete artifact.

Examples:
- a new prompt file entering the workspace
- a readme being written after a discussion
- a project folder getting a new canonical document because the system is ready for it

## Soil Surface

Immediate triage and first-contact processing.

Examples:
- the current open file
- the current conversation workspace
- recent drafts being checked, edited, or discarded
- live debugging work where context is being handled before it is sent downstream

## Creeks / Rivers

Structured retrieval and movement.

Examples:
- scripts that transform data
- notes that feed into a project document
- handoff packages
- search-and-read workflows
- reusable prompt files and skills

## Lakes / Reservoirs

Domain-organized storage.

Examples:
- `AnthroCybernetics/course-foundation-v2/`
- `Cascade/`
- `Documents/`
- `Projects/`
- `Skills/`

These are places where information is held intentionally, indexed by purpose.

## Groundwater

Deep memory and inherited structure.

Examples:
- durable AGENTS files
- old project decisions that still shape current work
- framework documents that keep reappearing across sessions
- hidden design assumptions in the way projects are built

The deepest groundwater layer in this workspace is the recurring architecture of:
- medulla / heartbeat
- ring / resonance / cycle logic
- tiered epistemics
- archive/tree / cup / kitchen imagery

## Return to Ocean

When a project is done well, it returns as reusable context:
- a summary
- a seed
- an archive artifact
- a new rule or prompt
- a pattern that can be applied elsewhere

This is the workspace learning from itself.

---

## Working claim

The workspace is healthiest when it has:
1. a clean ocean of raw material,
2. a strong evaporation layer,
3. active clouds of shared context,
4. trigger-based rain instead of forced schedule,
5. and a clean groundwater layer that stays unpolluted.