Repository navigation
Releases
Form & Force 0.2.2
[0.2.2] - 2026-11-16
The skill's tools keep one command-line contract and no longer pass a mistyped path in silence, and the component atlas is numbered and indexed. Around the skill, the repository is reorganised and gains an identity built from the maintainer's storyboard, a documentation site and complete project documents. The creative method is unchanged. No model evaluation has been run on this version.
Changed
-
Structures are called structures on the command line.
deal.mjs swap --part structureanddeal.mjs deck structuresnow match the documents and the printed hand.archetypeis still accepted, anddeck.json, history files and--jsonoutput keep the keyarchetype, so nothing stored changes. -
The component atlas numbers its sections 1 to 12 (they were 8.1 to 8.12, left over from an older document). It opens with an index of all 180 components by section, documents one-line cross-reference entries, and gives "Mobile-specific content hierarchy" the accessibility line it was missing.
floor.md§11 documents the scan's exit code for a path that does not exist. -
Repository tooling moved from
scripts/totooling/, so the repository no longer has twoscripts/folders in two languages. Commands change accordingly:python3 tooling/validate_repository.py,python3 tooling/package_skill.py,python3 tooling/release_check.py,python3 tooling/evals.py. -
A new identity, built from the maintainer's branding storyboard: an F made of two strokes, satin on dark grounds, with folds lit along their edges.
brand/holds the sources (the storyboard, the traced mark, colour and type tokens, the outlined wordmark), the guidelines, and every generated file:- marks and four lockups;
- app-icon tiles and favicons;
- README banners for light and dark, and the social preview;
- backgrounds;
- a brand board and a sheet of the variations explored.
.claude-plugin/icon.pngis re-rendered. -
The README is rewritten around the new documentation and brand.
-
Testing and releasing guides moved to
docs/development/. Testing now describes four kinds of verification: deterministic checks, browser checks, paid model evaluations and human judgement. -
The validator accepts a rendered brand file only when
brand/manifest.jsonrecords its exact hash. It lets the storyboard exceed the size limit so it is kept byte for byte, and it allows write permissions in one place only: the manual Pages workflow, while it stays manual. -
CI also checks the generated documentation tables and builds the documentation site, failing on any broken link or anchor.
Added
- Documentation:
- an introduction, a quick start, and a guide to how the method works;
- seven guides: creative direction, visual exploration, design language, motion, building, the floor and the jury, and reviewing existing work;
- reference pages for the tools, the reference library, the component atlas, the deck, templates, configuration and compatibility;
- troubleshooting, and development guides for extending the project and for the brand and site.
- A documentation site built from the Markdown by
site/build.py(Python standard library). It uses the brand's tokens and fonts, has light and dark themes, searches every page and every atlas entry, and fails the build on any broken link. A manual workflow deploys it to GitHub Pages. - Project documents:
ARCHITECTURE.md,ROADMAP.md,SUPPORT.md,CODE_OF_CONDUCT.md,CONTRIBUTORS.md,THIRD_PARTY_NOTICES.md, and issue templates for bugs, suggestions and design outcomes. tooling/sync_docs.pyregenerates documentation tables fromdeck.jsonand the atlas, with--checkfor CI.- Tests:
test_brand.py: contrast of the documentation's colours, brand provenance, current SVGs, and that the skill carries no brand;test_library.py: the atlas's format and index, and each reference's contents;test_site.py: the Markdown renderer and a full site build;- a command-line contract test for every tool.
Removed
assets/: the 0.1–0.2 identity (the banner, the social preview,mark.svgand the sheared disc rendered byassets/brand/render.py), replaced bybrand/. It remains in the history at the tagform-and-force--v0.2.1.
Fixed
- Source files named like a repository's documents are scanned.
slop_scan.mjsused to leave out any source file whose name starts with README, CHANGELOG, CONTRIBUTING, LICENSE, SECURITY, CODE_OF_CONDUCT, CLAUDE, AGENTS or SKILL, treating it like the repository's ownREADME.md. That skipped pages such aspages/security.tsx,components/Agents.jsxandChangelog.vuewithout a word. Now only Markdown files with those names are left out. - A path that is not a directory is an error, not an empty pass. Given a mistyped path,
slop_scan.mjsreported "0 files" and exited 0, even with--strict, so a CI gate could pass on a typo, andinspect_repo.mjscrashed. Both, andcontact_sheet.mjs, now say the path is not a directory and exit 2. - Every tool prints its usage with
--help, asSKILL.mdsays. Before this release:inspect_repo.mjsandslop_scan.mjsignored the flag and scanned the current directory;contact_sheet.mjsprinted usage only when run without arguments;deal.mjstreated-has an unknown command;motion_tokens.mjsprinted a short summary instead of its full usage notes;audit_page.mjsprinted its usage with the source code's comment markers left in.
audit_page.mjsrefuses an unknown option instead of auditing it as a URL.--viewport 390x844, a typo for--viewports, used to start an audit of a page called--viewport.deal.mjs deck <part>with an unknown part is an error instead of printing nothing.slop_scan.mjs's combined-sameness finding pointed toreferences/anti-slop-constitution.md, a file retired in 0.2.0. It now points tofloor.md§5 and §10.
Form & Force 0.2.1
Two fixes to the static scanner, slop_scan.mjs, both reported by @jcc2010 (#2, #3) and fixed in #5.
Update
claude plugin update form-and-force@form-and-forceNew installs: claude plugin marketplace add erenthedeveloper0/form-and-force, then claude plugin install form-and-force@form-and-force. For claude.ai or the Skills API, upload form-and-force-0.2.1.zip from the assets below.
Fixed
dead-linkno longer fails links built by string concatenation.'<a href="' + url + '">',"<a href='" + url + "'>"and Vue bindings such as:href="'/posts/' + slug"used to be reported as category C failures, because the two different quotes on either side of the concatenation were read as an empty href. The quotes around an empty or#value must now be the same character.href="#",href="",href='',href={"#"}, a bound:href="'#'"andjavascript:links are still caught, andhref={""}is now caught as well. (#2)- Server-rendered templates are no longer skipped without notice. A Laravel, Rails or Shopify project could come back as "0 files, no triggers found" and pass
--strictwithout a single view being read. Blade (.blade.php), ERB, EJS, Twig, Nunjucks (.njk), Jinja (.j2), Handlebars (.hbs), Mustache and Liquid files are now read in their own syntax: template comments are ignored, a=>or%>inside a tag does not end the tag, and a tag whose attributes are printed at render time is not reported as missing one. A template is read whatever it is called. The scan still leaves out a repository'sREADME.md,SECURITY.mdand similar documents, butsecurity.blade.phpis a page. (#3)
Changed
- Every scan states its coverage:
complete,partial,incompleteorempty.partial: templates were read as markup, or a source file was skipped for being over 800 kB or impossible to open.incomplete: some template files were not read.- Template formats the scan has no rules for (Pug, Haml, Slim, Razor, JSP, PHTML, HEEx, Go HTML) are counted and listed by extension.
- The JSON output gains a
coverageobject andsummary.coverage, andfilesScannednow counts only files that were opened.
slop_scan.mjs --strictexits 2 when there is no C finding but some template files were not read. It still exits 1 on any C finding and 0 otherwise. Without--strictit exits 0, as before.- Upgrade note: a project with templates may now report failures, or exit 2, where it used to exit 0. Those files were never checked before.
floor.md§11 documents coverage and the exit codes. Gate G1 now asks for a manual read of every file the scan did not read.- Known limits:
- A template is read as markup, so whatever a helper, a component or an include prints at render time (
link_to,image_tag, an attribute bag) is not seen. - Plain
.phpfiles are treated as code and are neither read nor listed.
- A template is read as markup, so whatever a helper, a component or an include prints at render time (
Verified for this release
- CI passes on Python 3.11, on pull request #5 and on
main. - All 145 offline tests pass. The 12 browser tests, which need Playwright, were skipped.
claude plugin validate --strict .passes on Claude Code 2.1.322.- The release gate (
release_check.py --online) reports READY: 33 pass, 0 fail. - In a clean Claude Code configuration, the plugin installs from GitHub and is enabled at 0.2.1.
- The ZIP is byte-for-byte reproducible and passes this repository's archive verifier.
Not yet verified
- Loading the plugin in a live Claude session, uploading the ZIP to claude.ai, and the Skills API.
- The behavioural evaluations have not been run on any release.
form-and-force-0.2.1.zip SHA-256: 386e07592428d0ca7693fa755b936dad85f9b0ed46c4e5a979dfbbd816989fbc
The full list of changes is in CHANGELOG.md.
Form & Force 0.2.0
[0.2.0] - 2026-11-11
The creative method is replaced. 0.1.0 produced coherent, careful and predictable work; this release changes how a direction is found, proved and judged. The diagnosis and the reasoning are in research/CREATIVE_METHOD.md. No model evaluation has been run on this version yet, so the improvement is argued and not measured.
Changed
SKILL.mdis rewritten around a new loop: read, set a register, write down the twin, deal a hand, sketch frames on a board, commit, derive the language from the chosen frame, build, clear the floor, sit the jury. It routes each kind of task to the references it needs and names what to skip.- Direction is dealt, drawn and then chosen.
deal.mjsdeals a lens, a compositional structure and a lead medium for each key frame, never two from one family, away from recent hands, and limited to material the project really has. Frames are sketched as standalone pages and compared on one board before any is preferred. - Ambition is set by a register (Quiet, Composed, Expressive, Radical) decided from the brief. Timid and overwrought are both failures, and restraint is no longer confused with either.
- Critique is a jury on rendered evidence. A cold read from the first screen, a test against the twin, ten dimensions marked in words with evidence, findings with a location and a change, and four verdicts, two of which send the work back to the idea. The Awwwards-weighted scorecard is gone.
- Hard gates are separate from judgement. Ten pass-or-fail gates and the registry of defaults live in
floor.md, which now also lists this skill's own attractor. - Motion values are generated from a character and stay inside documented bands. There are no example curves to copy.
audit_page.mjsalso writesfirst.jpg(the first screen at every width, with a blurred copy),flatplan.jpg(every screen in order) and design facts phrased as questions. Its evidence is composed piece by piece, so one failed image no longer discards the rest.project-modes.mdgives each mode its usual register, key frame and twin.component-atlas.md(renamed fromanti-slop-component-atlas.md) is positioned as a build-time reference, not a source of designs.- The report template follows the new method: direction and twin, the floor with actual results, the jury's marks and verdict, what is open.
Added
- References:
direction.md,composition.md,typography.md,colour-and-imagery.md,creative-coding.md,motion-direction.md,motion-craft.md,jury.md,floor.md. - Scripts:
deal.mjswithdeck.json(26 lenses in 8 families, 17 structures in 6 families, 8 lead media; no colours, typefaces or sizes),sketch_board.mjs,motion_tokens.mjs,motion_strip.mjs,design_facts.mjsand the sharedbrowser.mjs. motion_strip.mjsplays one moment in slow motion as a filmstrip and reports what animated, an interruption, the way back, layout shift, declared delays and a reduced-motion run.- Templates:
direction-brief.md,jury-brief.md(for a juror in a fresh context),motion-score.md. - Seven eval cases for the new method (direction before build, luxury that is not a palette, a radical site with a way out, a quiet site that is not bland, long-form editorial, a jury that rejects a competent twin, one engineered micro-interaction), with two fixtures. Written and structure-checked; not run.
- Regression tests for the deck, the dealer, the motion tokens, the design-facts thresholds and the agreement between documents and tools; browser tests for the rendered tools that run when
FF_PLAYWRIGHT_DIRis set and are skipped otherwise. - Research:
CREATIVE_METHOD.md,INSTRUMENT_CALIBRATION.md, and ledger section H.
Removed
creative-process.md,quality-gates.md,anti-slop-constitution.md,motion-intelligence.md,motion-implementation.md,visual-implementation.md, and thecreative-brief.mdandmotion-spec.mdtemplates. Their useful content moved into the references above.
Fixed
- The design-facts instrument was calibrated on 19 reference sites and 11 stock templates. The repetition measure no longer counts list rows, table rows or navigation as cards; two questions that flagged strong work and missed weak work were removed; a text-collision probe was added. Details and limits are in
research/INSTRUMENT_CALIBRATION.md.
Form & Force 0.1.0
The first public release of Form & Force: an Agent Skill and Claude Code plugin that carries web work from creative concept to verified code, covering creative direction, interface design and motion engineering.
Install
As a Claude Code plugin:
claude plugin marketplace add erenthedeveloper0/form-and-force
claude plugin install form-and-force@form-and-forceAs a standalone skill for claude.ai or the Skills API, upload form-and-force-0.1.0.zip from the assets below.
What's included
- The
form-and-forceskill. It inspects the project first, generates and scores divergent concepts, defines a project-specific design and motion language, builds in the existing stack, and reports only what actually ran. It comes with 15 references, 4 templates and 4 Node scripts:inspect_repo,slop_scan,audit_pageandcontact_sheet. - Free, offline checks. A repository validator, a reproducible ZIP builder, a release checklist and 80 tests, run by CI on every pull request. None of them needs an API key.
- An optional evaluation suite. Five cases that compare Claude with and without the plugin. Every run is billed, so it only runs when you start it yourself.
Verified for this release
claude plugin validate --strict .passes on Claude Code 2.1.308.- In a clean Claude Code configuration, the plugin installs from this repository and is enabled at 0.1.0.
- CI passes on Python 3.11.
- The skill ZIP passes this repository's archive verifier and the reference
agentskills validate(skills-ref 0.1.1).
Not yet verified
- Loading in a live Claude session, uploading the ZIP to claude.ai, and the Skills API.
- The behavioural evaluations have not been run, so there is no measured evidence yet that the skill improves results over Claude without it.
form-and-force-0.1.0.zip SHA-256: 8ac478d2b44d4cab1e8a1281fac3373bbd03966067ae9fd15676da0b868da4ee
The full list of changes is in CHANGELOG.md.