The repo outgrew a single-project layout: the reusable form-language template and the draft paper about the method were mixed in with Ecuador-specific ref/narratives/docs, and a second real project (Harbor House) had been growing inside templates/examples/ despite no longer being just a template demo. Split into: - method/ - the generalized form-language prompt template and the paper about the method, project-agnostic - projects/ecuador-museum/ - the original counter-proposal (ref/, narratives/, docs/, unchanged in content) - projects/harbor-house/ - promoted out of templates/examples/, now a peer project with its own ref/narratives/docs, including previously untracked pattern subset, narrative, and illustration material Root CLAUDE.md now covers the shared method and repo-wide conventions; each project gets its own CLAUDE.md with project-specific brief, inputs, and status. Updated all internal path references (paper.md, outline.md, per-project CLAUDE.md, settings.local.json). Dropped stray scratch files superseded by proper copies (all.txt, the already- converted pattern-language PDF). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
4.3 KiB
Harbor House — Travellers' Inn / Homeless-Relief Hostel, Sheffield
Paths below are relative to this directory (projects/harbor-house/)
unless stated otherwise. For the generalized method and repo-wide
conventions shared with other projects, see the root CLAUDE.md.
What this project is
Harbor House is a travellers' inn intended to relieve pressure on homeless
services, sited in Sheffield, South Yorkshire. Unlike the Ecuador museum
project (a public-advocacy counter-proposal responding to a real open
call), this project's origin brief is thin: "a travellers' inn to relieve
pressure on homeless services" in Sheffield, with scale and other
programmatic detail left to reasonable assumption rather than sourced from
an external brief document — see the "Note on method" at the top of
docs/sheffield.md for what was assumed versus sourced. It exists
primarily as the second validation project for this repo's generalized
method and form-language template (see root CLAUDE.md), proving the
pipeline generalizes beyond its Ecuador origin to an unrelated typology and
region.
Inputs (ref/)
harbor-house-pattern-language.txt— the project-specific pattern subset (Alexander/Ishikawa/Silverstein patterns such as TRAVELER'S INN, SLEEPING IN PUBLIC, COURTYARDS WHICH LIVE), the core design-constraint input for narrative writing, parallel to the Ecuador project'sA Pattern Language Ecuador National Museum.md.
Form language (docs/sheffield.md)
Generated using method/form-language-prompt-template.md — this is the
template's second worked example (the Ecuador form-language grammar,
projects/ecuador-museum/ref/Ecuadorian_Form_Language_Grammar.md, was the
first). Supplies the climate script for Sheffield, a form vocabulary drawn
from South Yorkshire/Pennine building history, composition syntax, and
color/ornament systems, to be combined with the pattern subset in the
narrative-writing prompt. Unlike the Ecuador worked example, this document
was generated with real web search available, so climate figures and
precedents are sourced inline rather than drawn from training-data
recollection alone — see its own "Note on method" section.
Deliverable
narratives/sheffield-narrative.md— the experiential design narrative for Harbor House, grounded in the pattern subset and form language above. Unlike Ecuador's four stakeholder viewpoints, this project currently has a single narrative, not split by stakeholder (visitor/staff/etc.) — no multi-viewpoint pass has been done here yet.docs/sheffield.docx— a pandoc export of the narrative for sharing (not committed to the repo, regeneratable, same convention as Ecuador'scounter-proposal.docx).docs/images/originals/— two ChatGPT-generated illustration contact sheets (9-panel storyboards walking through "a day at Harbor House": arrival, threshold, hearth, courtyard, gallery, shared meal, places apart, sleeping neighbourhoods, evening) grounded in the narrative.harbor-house-storyboard-draft.pngandharbor-house-storyboard.pngare two passes at the same 9-scene storyboard (the latter adds a title card and slightly revised captions); it's not confirmed which, if either, is meant to be final — unlike Ecuador, there's noillustration-prompts.mdequivalent recording the prompts that produced these, and no split into one image-per-scene at full resolution + compressed top-level copy (Ecuador'sdocs/images/convention).
Status
Less mature than the Ecuador project: pattern subset, form language, and one narrative exist, plus two draft illustration contact sheets. Not yet done, if this project is to reach parity with Ecuador's pipeline:
- Decide whether a single narrative is sufficient for this project's purpose, or whether multiple stakeholder viewpoints (e.g. resident, staff/caseworker, neighbor) are wanted.
- Reconcile the two storyboard images (pick one, or regenerate) and decide
whether to adopt Ecuador's per-scene illustration-prompt convention
(
docs/illustration-prompts.md+ one image per placeholder) instead of a single contact-sheet image. - No public-facing combined presentation document (Ecuador's
docs/counter-proposal.mdequivalent) exists yet — currently the narrative and images are separate deliverables, not merged into one.