The skill rewrite compressed two load-bearing instructions into ambient
context:
- "Navigate the browser to the URL" lost the signal that models with
Chrome MCP should proactively open the tab before the first poll.
Restored the forcing phrasing and the "before the first poll" anchor.
- "Claude Code can background the poll" read as permission rather than
prescription. Models fell back to foreground blocking by default.
Restructured harness guidance as a bulleted policy, prescriptive per
harness, with the reason attached (harness notifies on completion so
the conversation stays free).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Splits the skill into two register references (editorial, product),
replaces category-based theme selection with a forced physical-scene
inference, and introduces a four-step color strategy axis (Restrained /
Committed / Full palette / Drenched) with editorial permission for the
bold three.
Adds a seed mode to /impeccable document for pre-implementation
projects, updates /impeccable teach Step 5 to offer the seed path, and
grows /impeccable shape with Design Direction + Scope intake
(fidelity, breadth, interactivity, time). Extends live-mode variant
distinctness to forbid three variants sharing theme and dominant hue.
Also drops the anti-pattern validator coupling, consolidates a11y into
audit.md, and updates CLAUDE.md with the register architecture.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds a full annotation pipeline to /impeccable live. On Go, the browser
captures the selected element as a PNG (with annotations composed in),
uploads it to the live helper, and sends the generate event with the
screenshot path so the agent reads user intent visually instead of from
HTML alone.
Annotation tools (while an element is picked):
- Click inside the outline to drop a magenta comment pin with a text input
- Drag to paint a magenta SVG stroke (5 px click-vs-drag threshold)
- Click a pin to edit; double-click to delete; drag a pin to reposition
- Click a stroke to delete it (wider invisible hit path)
- Clear chip top-right wipes everything; hidden when no annotations
Capture pipeline:
- modern-screenshot vendored as an IIFE (scripts/modern-screenshot.umd.js)
and lazy-loaded from the live helper
- Font fix: cross-origin @font-face rules are fetched and fonts are inlined
as base64 data URIs before being handed to modern-screenshot via
font.cssText, since SVGs rasterized via canvas can't fetch external
resources (fix for "Impeccable" rendering bold-serif and items wrapping
wrong in the capture)
- Annotations are temporarily attached to the live element (not only the
clone) so computed styles resolve during the embed pass
- Session screenshots live in .impeccable-live/annotations/session-*/ in
the project root (gitignored) so the agent's Read tool doesn't trip a
per-path permission prompt
Loading shader (activates during GENERATING):
- WebGL overlay rendering the captured PNG as a halftone — cells with
luma-driven dot radius, rendered on paper-cream underneath a magenta
roller that sweeps top-to-bottom with a 3.4s cycle and clean overshoot
- Fixed asymmetric bandAt() using one-sided smoothsteps (previous reversed
smoothstep was undefined on d>0, giving "trail=1 everywhere below")
- Graceful <img> fallback when WebGL is unavailable; prefers-reduced-motion
freezes the band at t=0
Server:
- POST /annotation endpoint (raw image/png body, token + eventId query),
session-scoped tmpdir cleaned up on shutdown
- GET /modern-screenshot.js serves the vendored UMD with aggressive caching
- Optional screenshotPath / comments / strokes fields on generate events
- Fixed pre-existing /source crash on ENOENT (writeHead called twice)
Agent side:
- reference/live.md step 0 tells the agent to Read the screenshot first,
with four rules for interpreting annotations: comments are position-
anchored and scoped to the sub-element under their {x,y}; strokes are
gestures (loop=focus, arrow=direction, cross=delete); comments and
strokes are independent unless adjacent; don't silently guess on
ambiguous strokes
Also:
- Generating bar no longer claims "Generating 1 of 3..." (variants arrive
atomically) — now says "Generating N variants..."
- tests/live-server.test.mjs fixed to read the PID file from project root,
matching the server; adds coverage for the new endpoints and validator
fields
- .impeccable-live/ added to .gitignore
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
live-server.mjs stop now runs live-inject.mjs --remove after the HTTP
server shuts down, so HTML entries do not keep loading a dead localhost
live.js URL. Add stop --keep-inject to stop only the helper.
Update reference/live.md cleanup steps and sync all provider skill copies.
Made-with: Cursor
Document that Cursor Composer should run live-poll blocking in the same turn
(with a link to Cursor subagent foreground vs background docs). Claude Code may
use a background poll when the harness surfaces completion. Sync reference to
all provider skill bundles.
Made-with: Cursor
Global bar is now a single compact unit: Pick → Detect → DESIGN.md
with the Impeccable brand mark as a full-height slab on the left.
Labels are icon-only at rest and expand as a group when the bar is
hovered, so moving the cursor across buttons no longer triggers
per-button layout thrashing. Button styling flips automatically based
on the page's ambient luminance — dark bar on light pages, light bar
on dark pages — so the bar doesn't fight with the host design.
The DESIGN.md panel chrome now matches the bar (same surface, hairline,
mono filename title) while the body canvas stays neutral so tile colors
and rendered component primitives look true. The separate floating
"Design" FAB is gone; the panel toggle lives in the bar. Panel always
starts closed; only the tab and collapsed-section preferences persist.
Other fixes: toggling pick off now clears the selection + hides the
contextual bar and action picker. PRODUCT.md presence is the signal
for "project context loaded" for variant generation (DESIGN.md lives
under its own empty-state in the panel). The live-poll client sets a
global undici dispatcher with no headers/body timeouts so one poll can
sit open indefinitely — fixes the silent "fetch failed" that killed
polls at the 5-minute mark.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The live-mode float-bar now includes a "Design" toggle that slides a
panel in from the right. Tile-based layout (neutral canvas, one level
of hierarchy, no nested cards) with color swatches and tonal ramps,
typography specimens, corner radii, shadow previews, and *live
component primitives* rendered from the project's real tokens.
Collapsible Named Rules / Do's-and-Don'ts / Overview hold the
narrative context without crowding the tiles.
The /impeccable document command now writes a DESIGN.json sidecar
alongside DESIGN.md. The sidecar carries structured tokens plus
self-contained HTML+CSS snippets per component — this is what lets
the panel render each project's actual button/input/nav instead of
generic approximations. The document spec documents the translation
rules for Tailwind, CSS-in-JS, shadcn, and framework components.
The live server exposes /design-system.json and /design-system/raw.
If DESIGN.json is missing but DESIGN.md is present, the panel falls
back to a limited "basic view" parsed from the markdown and prompts
the user to run /impeccable document for the full visualization. A
stale hint appears when DESIGN.md has been edited after DESIGN.json.
Includes a hand-authored DESIGN.json for this project so the panel
has something to render against out of the box, and a deterministic
DESIGN.md parser (design-parser.mjs) as the fallback source.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
typography.md had a parallel 4-step font-selection procedure and a smaller
banned-fonts list (5 fonts: Inter, Roboto, Open Sans, Lato, Montserrat)
that duplicated SKILL.md's authoritative <font_selection_procedure> with
its 23-font list. Removed the duplicate procedure and deferred to SKILL.md
for the banned list. Kept the unique material: anti-reflex corrections,
system-font note, pairing principles, web font loading, OpenType, fluid
type guidance, accessibility — and all of the scale/rhythm/measure
content that SKILL.md doesn't cover.
pin.md removed for the same reason as context.md: SKILL.md's inlined
pin section already covers what an agent needs (what pin does, usage,
valid commands, how to report back). No value in the indirection.
SKILL.md: 386 → 388 lines (slight growth from inlining pin details)
typography.md: 142 → 132 lines
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The main skill now has enough detail to stand alone: two files, load
command, no-truncate rule, never-infer warning, session cache, teach
fallback, DESIGN.md nudge, and the live-mode "already warmed"
exception. context.md was indirection without added value.
Inlined the two bits from context.md worth keeping:
- Content validity: treat empty / <200 chars / [TODO]-placeholder
PRODUCT.md as missing
- live.mjs auto-warms, don't double-load with load-context.mjs
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Before: 422 lines. After: 386 lines. The cut is conservative because
evals/AGENT.md revealed that most of what looked like bloat is actually
load-bearing: the font_selection_procedure with its 23-font ban list
(lesson 6), the theme_selection with audience examples (lesson 9), and
the absolute_bans with literal CSS patterns (lesson 7) all drive
measurable eval improvements and must stay inline.
What moved out of SKILL.md:
- Context Gathering Protocol (52 → ~18 lines). The full protocol — cache
semantics, dispatch tree, teach/document/live exceptions, why-it-matters
— moved to reference/context.md. SKILL.md keeps only the compact hook:
load command, "never infer from codebase" warning, and pointer.
- Pin/Unpin (14 → 6 lines). Details moved to reference/pin.md.
- Spatial principles: dropped 4pt-vs-8pt rationale, gap-vs-margins CSS
technicality, and container-queries-vs-viewport explanation (not
load-bearing in the main skill). Kept all load-bearing rules
including the 80-char body-text line (detector-backed).
What did NOT move (load-bearing per evals):
- <font_selection_procedure> with the 23-font ban list
- <theme_selection> with the 8 audience examples
- <absolute_bans> with literal CSS patterns
- All XML tag structure (lesson 8: XML works better than markdown
for reasoning models, especially OpenAI)
Also added:
- reference/context.md (new) — full context protocol
- reference/pin.md (new) — full pin/unpin docs
- "Never infer brand, audience, or tone from the codebase" warning
restored to SKILL.md (was dropped in an earlier refactor)
Fixed:
- reference/colorize.md had "Accent borders: Add colored left/top
borders to cards or sections" which directly contradicted
SKILL.md's absolute_ban on border-left/right > 1px. Rewrote the
accent-border advice to use hairline borders, surface tints, or
leading glyphs instead, with an explicit reference to the ban.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two observed failure modes:
1. Smoke-test style truncation (`| head -N`) in bash commands defeats the
whole point of load-context.mjs — Claude needs the FULL file contents,
not the first few lines of JSON.
2. The old protocol didn't clearly explain session caching, leading to
repeated load-context.mjs calls across commands in the same session
(thousands of wasted tokens on 3-5KB files re-fetched 3-5 times).
Context Gathering Protocol rewrite:
- PRODUCT.md required (blocker), DESIGN.md optional (one-line nudge if
missing). Greenfield projects can't yet have a DESIGN.md to document.
- Explicit session cache: if content is in conversation history, do not
re-fetch. Exceptions listed (after teach/document/manual edit).
- Explicit "never truncate" rule: consume the full load-context.mjs
output, never pipe through head/tail/grep/jq with field filters.
- Content validity check: hasProduct=true but content <200 chars or
full of [TODO] markers = treat as missing, run teach.
- Missing-PRODUCT.md flow spells out task resumption: user asked for
/impeccable polish ButtonGroup, we must run teach, then RESUME polish
of ButtonGroup with fresh context — not silently abandon intent.
- Three explicit exceptions to the protocol:
- /impeccable teach skips it (teach creates PRODUCT.md)
- /impeccable document loads PRODUCT.md only (creates DESIGN.md)
- /impeccable live already warms context via live.mjs — don't also
run load-context.mjs
teach.md Step 6 and document.md Step 5 now re-run load-context.mjs at
the end so the freshly-written files surface in conversation history
and subsequent commands use the new version, not a stale earlier read.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The format at https://stitch.withgoogle.com/docs/design-md/format/ defines
exactly six sections in a fixed order: Overview, Colors, Typography,
Elevation, Components, Do's and Don'ts. Our previous format used
non-compliant names (Visual Theme & Atmosphere, Color Palette & Roles,
Typography Rules, Component Stylings), a non-existent "Layout Principles"
section, and had no Do's and Don'ts.
Changes to reference/document.md:
- Fixed section list to match the spec character-for-character
- Added mandatory "Creative North Star" pattern at top of Overview
- Added Named Rules pattern (e.g. "The No-Line Rule") — stickier than
bullet lists for AI consumers, mirrors Stitch's own generator output
- Added explicit Do's and Don'ts section with concrete, forceful guardrails
- Elevation is now its own section (was buried in Components)
- Layout/motion/responsive content folds into Overview + Components
rather than inventing new top-level sections
- Guidance on forceful voice ("prohibited"/"forbidden"/"never") matching
PRODUCT.md's expert-decisive tone
- Pitfalls section warns against renaming sections or adding new ones
Changes to our DESIGN.md:
- Rewrote to use spec-compliant section headers with evocative subtitles
(e.g. "## 2. Colors: The Warm-Paper Palette")
- Opened with "Creative North Star: The Editorial Sanctuary"
- Added 11 Named Rules across sections (The One Voice Rule, The Paper-
Not-White Rule, The OKLCH-Only Rule, The Italic-Is-Voice Rule,
The 1.6 Leading Rule, The Fluid-Headlines-Only Rule, The Flat-By-
Default Rule, The Low-Alpha Rule, The Tinted-Shadow-Only-For-Accent
Rule, plus the existing implicit ones)
- Full Do's and Don'ts section with 10 Dos and 15 Don'ts, many of which
translate PRODUCT.md anti-references into concrete prohibitions
- Elevation section documents shadow vocabulary separately from Components
- Layout/spacing/motion content relocated to a sub-section under Components
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Pioneers a two-file convention for project context:
- PRODUCT.md (strategic): users, brand, principles — answers who/what/why
- DESIGN.md (visual): follows Google's Stitch DESIGN.md spec — answers how-it-looks
Both files live at the repo root. Filename matching is case-insensitive.
DESIGN.md wins on visual conflicts, PRODUCT.md wins on strategic/voice.
Legacy .impeccable.md is auto-migrated to PRODUCT.md on first read by the
new shared loader. This is silent and one-shot — the rename is permanent.
What changed:
- New scripts/load-context.mjs: shared context loader used by every command
that needs project context. Reads both files, handles legacy migration.
- New reference/document.md: /impeccable document command that generates
DESIGN.md by auto-extracting tokens (colors, typography, spacing, radii,
shadows, components) from CSS/Tailwind/theme files, then asking the user
to confirm descriptive language for atmosphere and color character.
Follows Google's Stitch DESIGN.md format for tool compatibility.
- SKILL.md Context Gathering Protocol updated to load both files and
nudge the user to run /impeccable document when DESIGN.md is missing.
- reference/teach.md rewritten to split discovery cleanly: strategic
questions go to PRODUCT.md, visual/design-system work is delegated to
/impeccable document (skipped on empty projects).
- reference/live.md consumes {product, design, productPath, designPath,
migrated} from the loader instead of a single context blob.
- scripts/live.mjs uses the shared loader instead of inline file reading.
- Command count updated 22 → 23 (new: document). Metadata, router table,
command menu, periodic table viz, and homepage data all updated.
- .gitignore adds PRODUCT.md + DESIGN.md (repo-local, not shared).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Previously only bolder/quieter/animate/colorize/typeset/layout had
variant diversity rules. Added the same level of guidance for distill,
polish, adapt, delight, and overdrive so every live action has a
clear "each variant must differ on THIS axis" rule.
Also noted that overdrive should skip its reference's "propose and ask"
step in live mode (it's non-interactive — the user picks from variants).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two failure modes observed:
1. Claude generates N near-identical variants (small shade/size tweaks)
instead of meaningfully different design directions
2. When a sub-command like /bolder is chosen in the picker, Claude skips
loading reference/bolder.md and generates generic variants
Fixes:
- "Load reference file" is now a MANDATORY Step 2a, separate and
non-negotiable, called out as a critical failure to skip
- Added Step 2b "Plan 3+ distinctly different directions" with 7
structural axes variants must differ on (hierarchy, layout topology,
typography system, color strategy, density, tone, decomposition)
- Added action-specific diversity rules (bolder = different dimensions,
animate = different motion vocabulary, colorize = different hues, etc.)
- Freeform prompt guidance: honor the prompt direction but explore
meaningfully different interpretations, not three near-copies
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Previously, starting live mode required ~5-6 sequential bash calls:
read .impeccable.md, start server, check config, read reference, inject
tag, verify. The new live.mjs does all of this in a single command
(~340ms cold, ~90ms when reusing a running server) and returns everything
the agent needs in one JSON blob.
Workflow is now:
1. node live.mjs # start + inject + load context (1 bash call)
2. navigate browser # optional MCP call
3. node live-poll.mjs # enter poll loop
Reference doc collapsed to a single "Start Live Mode" section with the
one-command path plus a first-time config creation fallback.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
First live run: agent auto-detects framework and writes a small config.json
(file, insertBefore/insertAfter anchor, comment syntax). Every subsequent
run: live-inject.mjs handles insert/remove deterministically, no LLM needed.
The config lives at {scripts_path}/config.json and is gitignored — it's a
per-project cache that wipes on skill update and regenerates on next use.
- New live-inject.mjs: --port (insert), --remove, --check modes
- Idempotent insert: re-running with a different port replaces cleanly
- Reference doc: one-time detection step, then instant insert/remove
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Arrow keys now pass through to element picker when the freeform input
is empty, instead of being swallowed by stopPropagation
- Arrow nav works in both PICKING and CONFIGURING states, so you can
change your element selection while the config bar is open
- Clicking outside the selected element and bar returns to PICKING mode,
matching the expected deselect behavior
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Accept and discard in live variant mode are now handled by a deterministic
script (live-accept.mjs) that runs inside the poller before returning to
the agent. The browser updates the DOM instantly on click (fire-and-forget)
so the user is never blocked waiting for LLM-driven file cleanup.
Key changes:
- New live-accept.mjs: deterministic accept/discard file operations
- Poller auto-runs accept script for accept/discard events (_acceptResult)
- Browser handleAccept() now commits DOM change instantly, no SAVING state
- CSS+HTML colocated in one write (style tag inside variant wrapper)
- SSE heartbeat every 30s prevents silent connection drops
- Poll timeout increased from 2min to 10min
- EventSource onopen resets retry counter for reliable reconnection
- Server --background flag for clean single-command startup
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Comprehensive architecture decision record covering the live variant
mode: context, key decisions (source modification over DOM patching,
SSE over WebSocket, self-contained skill scripts, HTTP long-poll for
agent), full architecture diagram with message flows, variant wrapper
format, browser UI states, session persistence, security model,
server resilience, performance optimizations, test coverage, known
limitations, and future work.
Also picks up improvements from parallel thread: poll timeout bumped
to 10 min, SSE heartbeat every 30s, and other minor fixes.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
os.tmpdir() returns /var/folders/.../T/ on macOS, not /tmp/. The skill
reference was telling the agent to cat /tmp/impeccable-live.json which
didn't exist. Moving the PID file to the project root makes it
predictable across platforms and project-scoped (multiple projects can
run independent live sessions).
Changed in: live-server.mjs, live-poll.mjs, live.md reference.
Added .impeccable-live.json to .gitignore.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The server is started with & (backgrounded), so its stdout output isn't
captured by the agent's Bash tool. The skill reference now tells the
agent to sleep 2s then cat the PID file (/tmp/impeccable-live.json) to
get the port and token. The PID file is written by the server as soon
as it starts listening.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Compact floating pill at the bottom center of the viewport, always
visible during live mode. Matches the action bar's light, translucent
aesthetic with brand-tinted active states.
Controls:
- "Impeccable" brand mark (capitalized, brand magenta)
- Detect toggle: eye icon, loads anti-pattern scanner in extension mode,
waits for impeccable-ready before first scan, shows issue count badge
inside the button. Toggle off removes overlays.
- Pick toggle: crosshair icon, enables/disables element picker. Active
by default. When pick is active, detect overlays get pointer-events:
none so the picker sees through them.
- Exit button: sends exit event and tears down all UI.
Detect + pick coexistence fixes:
- Picker highlight z-index raised above detect overlays (100001 vs 99999)
so the selection outline and element path are always visible.
- Removed layout-property transitions (top/left/width/height) from the
highlight to avoid triggering the anti-pattern detector and to give
instant cursor tracking.
- First-click-on-detect fix: script loads async, scan command is queued
until the impeccable-ready postMessage arrives.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1. Skill reference: poll should run as background task with no timeout.
Changed "blocking poll loop" to "background task, no timeout" so the
agent keeps the main conversation free for other work.
2. Resume restores selectedAction from localStorage: the bar was showing
"Freeform" after page reload even when the user picked "Bolder". Also
improved selectedElement targeting to prefer the visible variant's
content over the wrapper parent.
3. Discard no longer shows "Applying variant...": accept shows the
saving→confirmed flow, but discard now dismisses immediately and
cleans up the DOM. Different intent, different UX.
4. Picker works after discard: cleanup() now removes the variant wrapper
from the live DOM and restores the original element. Previously the
stale wrapper with data-impeccable-variant attributes confused the
picker's isPickable/own checks.
5. Stop live mode: added "Stopping Live Mode" section to the skill
reference. The user can say "stop live mode" in the conversation, and
the agent proceeds to cleanup (remove script tag, stop server).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The static "[command] [target]" hint didn't help users discover available
commands. The build now reads command-metadata.json and groups commands
by category (from SKILL_CATEGORIES) with middle-dot separators for
natural line-breaking in the prompt bar.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Tests:
- tests/live-wrap.test.mjs (26 tests): unit tests for buildSearchQueries,
findElement, findClosingLine, detectCommentSyntax (20 pure function
tests) + integration tests for the full wrapCli on HTML and JSX
fixtures with temp dirs (6 tests covering wrapping, ID/class lookup,
error handling, content preservation).
- tests/live-server.test.mjs (15 tests): integration tests that start a
real server on port 8499, then test /health, /live.js, /detect.js,
/poll (timeout + auth), /events POST (validation + auth), browser→agent
event flow (POST event → poll receives it), agent→browser SSE flow
(POST reply → SSE stream delivers it), /source (read, path traversal
rejection, auth, 404).
Also:
- Added auto-execute guards to live-wrap.mjs and live-poll.mjs so they
work when run directly with `node live-wrap.mjs ...` (needed for both
skill instructions and integration tests).
- Exported buildSearchQueries, findElement, findClosingLine,
detectCommentSyntax from live-wrap.mjs for unit testing.
- Updated package.json test script to include the new test files.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1. CLI cleanup: removed live, poll, and wrap commands from bin/cli.js
and the liveCli export from detect-antipatterns.mjs. These now live
exclusively in the skill scripts (node scripts_path/live-server.mjs).
2. Deleted src/live/: server.mjs, poll.mjs, wrap.mjs, browser.js,
protocol.mjs. The source of truth is now source/skills/impeccable/
scripts/live-*.
3. Graceful server-lost handling: the browser tracks SSE reconnection
attempts (max 5). After exhausting retries, it cleans up the UI:
hides the bar, highlight, and cycler, shows a "Live server
disconnected" toast, resets state to IDLE. This handles agent
crashes, server kills, and network issues without leaving the
browser stuck in a "Generating..." state.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two architectural changes that make the live variant mode self-contained:
1. SSE replaces WebSocket: the server now uses Server-Sent Events for
server→browser push and regular fetch POST for browser→server
events. This eliminates the ws npm dependency entirely. The live
server is now zero-dependency pure Node.js (http, crypto, fs, net).
Browser: EventSource replaces WebSocket. sendEvent() uses fetch POST.
Server: GET /events returns SSE stream, POST /events receives browser
events. All other endpoints (poll, source, health, stop) unchanged.
2. Scripts moved to source/skills/impeccable/scripts/: live-server.mjs,
live-poll.mjs, live-wrap.mjs, live-browser.js are now part of the
skill itself. Users who install the skill via npx skills get the live
mode without needing npm install impeccable separately.
The skill reference uses {{scripts_path}}/live-server.mjs etc.
The CLI (bin/cli.js) delegates to the skill scripts as a convenience.
Removed ws from package.json dependencies.
The old src/live/ files remain as the development copy. The build system
syncs source/skills/ to all harness dirs (11 providers).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
For dev servers without HMR (Bun static imports, simple HTTP servers),
the browser can't see file changes automatically. Three changes fix this:
1. /source endpoint on live server: reads a project file from disk,
gated by session token + path-traversal guard. The browser fetches
the raw HTML directly, bypassing the dev server's cache.
2. poll --reply --file flag: agent passes the source file path when
replying done. The browser receives it via WS and knows where to
fetch. Skill reference updated to always include --file.
3. Browser injectVariantsFromSource(): on "done" with 0 DOM variants,
fetches the raw HTML from /source, parses with DOMParser, extracts
the variant wrapper, finds the matching element in the live DOM by
class/ID, and replaces it. MutationObserver picks up the injected
variants and the cycling bar appears.
Also: wrap CLI no longer hides the original element (was display:none).
The original stays visible until the first variant arrives, preventing
a flash of empty content between wrap and variant insertion.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Three optimizations to cut the generate loop from ~40s to ~15-20s:
1. wrap CLI helper (src/live/wrap.mjs): finds an element in source
by ID, class names, or tag+class combo, wraps it in the variant
container with original snapshot, and returns the file path + insert
line. Replaces 3-4 agent tool calls (grep + read + edit) with one.
Supports --element-id, --classes (comma-separated), --tag, --query
(fallback). Searches in priority order: ID > class combo > single
class > raw text. Auto-detects comment syntax (HTML vs JSX).
2. Batch variant writes: skill reference updated to instruct the agent
to write ALL variants in a single file edit instead of one per
variant. Saves N-1 tool call round-trips (~3-5s each).
3. Page URL in generate event: browser now includes location.pathname
so the agent can map URL to source file directly (/ = index.html,
/about = about.tsx, etc.) without grepping.
Net effect: agent flow is now 4 tool calls (wrap + edit + read-variant
+ poll-reply) instead of 8+ (grep + read + create-wrapper + N edits
+ poll-reply).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
New feature: /impeccable live starts an interactive visual iteration server.
Users select elements in the browser, pick a design action (bolder, quieter,
etc.), and the agent generates HTML+CSS variants written directly to source.
The dev server's HMR hot-swaps them in, and MutationObserver progressively
reveals each variant in a cycler UI as it arrives.
Architecture:
- src/live/server.mjs: HTTP + WebSocket server with session token auth,
long-poll /poll endpoint for the agent, WebSocket for the browser
- src/live/poll.mjs: CLI client (npx impeccable poll / poll --reply)
- src/live/browser.js: element picker with keyboard nav (arrows=siblings,
shift+arrows=parent/child), action panel (12 commands, freeform input,
variant count), variant cycler with progressive reveal via MutationObserver
- src/live/protocol.mjs: shared message types and event validation
- source/skills/impeccable/reference/live.md: agent loop instructions
(inject script, poll loop, generate variants, accept/discard, cleanup)
CLI changes:
- bin/cli.js: added "poll" top-level command
- src/detect-antipatterns.mjs: liveCli() now delegates to src/live/server.mjs
- package.json: added ws dependency
Registered /impeccable live as command #22 across all standard locations.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Pre-3.0, onboard was folded into /harden when we were trying to reduce
namespace pollution. In the single-skill model that tradeoff is gone,
so the weakest of the old merges is the first to undo.
Harden and onboard live in different mental modes. Harden is defensive
(edge cases, i18n, overflow, errors). Onboard is activation (first-run
flows, empty states as CTAs, progressive disclosure). A user thinking
"design the onboarding flow" was never going to type /impeccable harden.
Changes:
- New reference file at source/skills/impeccable/reference/onboard.md,
restored from the pre-merge version in git history rather than the
condensed 33-line summary that was in harden.md.
- Removed the "Onboarding & First-Run Experience" section from
source/skills/impeccable/reference/harden.md.
- Updated harden description/editorial/process-steps to drop onboarding
keywords; split commandProcessSteps so harden stays focused on
production resilience and onboard gets its own phases.
- Registered onboard in: SKILL.md description + command menu + router
table, command-metadata.json, IMPECCABLE_SUB_COMMANDS, pin.mjs
VALID_COMMANDS, SKILL_CATEGORIES, COMMAND_RELATIONSHIPS, data.js
commandCategories + commandProcessSteps + commandRelationships,
framework-viz commandSymbols + commandNumbers.
- Reused the existing content/site/skills/onboard.md editorial wrapper
(it was orphaned by the merge but never deleted), updating it to use
/impeccable onboard.
- Bumped all user-facing count references 20 -> 21: public/index.html,
CLAUDE.md, README.md, NOTICE.md, plugin.json, marketplace.json,
sitemap.xml, build-sub-pages.js.
- Harness dir audit.md and critique.md changes are the
{{available_commands}} placeholder expanding to include onboard.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Biggest change in a while. Users previously had 18 standalone skill
entries cluttering their /menu; now they have one entry (/impeccable)
that routes to 20 specialized commands via argument dispatch. The pin
mechanism (/impeccable pin audit) restores standalone shortcuts on
demand for commands users hit all the time.
## Architecture
- Single /impeccable skill with command router section in SKILL.md
- 20 commands served via reference files under source/skills/impeccable/reference/
- /impeccable pin <command> creates a lightweight redirect shim so users
who prefer /audit, /polish, etc. can still have them
- Context gathering (teach) auto-runs on first use
- command-metadata.json is the single source of truth for command
descriptions, argument hints, and relationships
## Site rewrite
- Docs URL: /skills renamed to /docs (with /skills permanent redirects)
- Homepage hero frames Impeccable as "one skill with 20 commands"
- "Get Started" split into 50/50 install + how-to-use with editorial
numbered steps, /impeccable shown as the home command with three modes
- New /docs overview: home command hero card + dense category rows
matching the old cheatsheet density, with leads-to/pairs-with/
combines-with relationship metadata served from a shared source
- Cheatsheet merged into /docs, /cheatsheet redirects
- Magazine spread and mobile cards show /impeccable as a stacked
namespace label above the command name at full display size
- Periodic table updated with craft/teach/extract as first-class cells
- Skill detail pages generate from reference files, with an editorial
wrapper per command for tagline + body
- Tutorials and anti-patterns pages updated to use /impeccable <cmd>
## Build system
- Dead code removed (scripts/lib/transformers/shared.js)
- Build log wording fixed ("1 skill" not "1 skills (1 user-invocable)")
- generateApiData fallback branch removed (throws loudly if metadata
missing instead of silently degrading)
- Commands API includes editorial tagline alongside the long description;
UI surfaces prefer tagline for human display, description for auto-
trigger keyword matching
## Gitignore
- Added .claude/scheduled_tasks.lock, .claude/settings.local.json to
ignore list (local Claude Code state that should not be tracked).
- Harness skill directories (.claude/skills/, .agents/skills/, etc.)
remain tracked by design: npx skills reads them from this repo at
install time and they enable clean submodule use.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The post-update-cleanup section now instructs the AI to tell the user
what's happening and why before running the script, so file deletions
don't feel unexpected.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Delete source/skills/ directories for deprecated skills (arrange,
normalize, onboard, extract, frontend-design, teach-impeccable).
The cleanup script handles migration; stubs are no longer needed.
- Add "npx skills update" command to the Stay Updated install section
- Rewrite FAQ update answer: lead with npx skills update, add
troubleshooting for failed updates (re-install + run /impeccable)
- Run cleanup script in `npx impeccable skills update` before
delegating to npx skills update, preventing failures from
deprecated entries in skills-lock.json
- Run cleanup script after `npx impeccable skills install` to remove
leftover files from previous versions
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Bundled as source/skills/impeccable/scripts/cleanup-deprecated.mjs,
runs via the self-deleting <post-update-cleanup> section in the skill.
The script:
- Finds all harness skill dirs (.claude, .cursor, .agents, etc.)
- Deletes deprecated skill directories (arrange, normalize, onboard,
extract, frontend-design, teach-impeccable) and i-prefixed variants
- Verifies each file contains "impeccable" before deleting to avoid
touching unrelated user skills with the same name
- Handles both symlinks and regular directories
- Removes matching entries from skills-lock.json (only if source is
pbakaus/impeccable)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The build counts active commands by checking for DEPRECATED in the
skill description. Without this, the old source files were still
counted as active (22) while the site says 18, failing CI.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Version bump across package.json, plugin.json, marketplace.json
- Changelog entry for v2.1 in index.html
- Hero version link updated
- Added <post-update-cleanup> section to impeccable SKILL.md that
detects and removes leftover files from renamed/merged skills
(arrange, normalize, onboard, extract, frontend-design,
teach-impeccable). Verifies files contain "impeccable" before
deleting to avoid touching unrelated user skills. Self-deletes
after first run so it only executes once per update.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Syne stays in reflex_fonts_to_reject but no longer gets a standalone
DON'T rule in the skill or a separate LLM-only entry in the catalog.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Clarify that /impeccable craft runs /shape internally (not the other way around)
- Add three-mode documentation (freeform/craft/teach) to /impeccable page with anchor links
- Add sidebar sub-links for /impeccable craft and /impeccable teach
- Fix hallucinated npx impeccable live description in tutorial and visual-mode page
- Remove nonsensical "Do not skip the independent part" from critique tutorial
- Make Step 4 less prescriptive (users can fix all at once or one-by-one)
- Improve CLI and browser extension install copy with specific features and use cases
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The skill's "don't use Inter / don't use dark / don't center" negatives
were creating new attractors (the model picks Fraunces / light / grid
instead, every time). Inline always-applicable principles into SKILL.md,
add a font selection anti-attractor procedure that forces the model to
enumerate AND reject its reflex defaults, switch high-stakes blocks to
XML structure, tighten side-tab and gradient-text bans to specific CSS
patterns, ban Syne explicitly, and strip named font/color prescriptions
from the references. Validated against the internal eval framework on
Qwen 3.6 Plus across 7 niches: Fraunces dropped from 92% to 0% on kids
reading, side-tabs from 76% to 20% on vintage moto, no theme regressions.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
A new icon-tile-stack detection (the canonical AI feature-card with a
small rounded-square icon container above a heading), backed by a
two-column TDD fixture, plus a single-source-of-truth design that ties
the engine to the impeccable skill so they can no longer drift silently.
Detection
- New icon-tile-stack rule (slop): heading's previousElementSibling is
a 32–128px rounded-square element with a non-transparent background
or border, contains an svg/icon-i child, and sits above (not next to)
the heading. Excludes round avatars, wide thumbnails, side-by-side
layouts, tiny icons, and hero images.
- Two-column fixture convention: a single icon-tile-stack.html with a
flag column (4 cases) and pass column (6 cases), with snippet-text
matching used by the fixture test.
Single source of truth
- Each ANTIPATTERNS entry can now declare skillSection + skillGuideline.
18 of 25 rules carry these fields; the build's new
validateAntipatternRules() in scripts/build.js fails if any declared
skillGuideline isn't found verbatim in the right SKILL.md section.
- scripts/build-extension.js now includes the description field in
extension/detector/antipatterns.json (it was previously dropped).
- The existing count validator was promoted from warn to error so
command count drift fails the build the same way detection drift does.
Impeccable skill DON'Ts
- Added 4 new top-level DON'Ts that target real default AI behavior:
single-font, flat-type-hierarchy, all-caps-body, line-length.
- Cut 7 new DON'Ts I had drafted (tight-leading, tiny-text, wide-tracking,
justified-text, low-contrast, cramped-padding, skipped-heading) because
they teach things every model already knows from CSS/a11y basics. The
detector still catches all of them.
Stale count cleanup
- 22 commands → 21 across 17 references in HTML, README, NOTICE, AGENTS,
plugin.json, marketplace.json (left over from the validate skill removal).
- Dropped the hand-coded "212 design guidelines" marketing copy on the
homepage, which never mapped to any real count.
Sub-agent
- New private .claude/agents/anti-patterns.md captures the full TDD
recipe, schema, plug-in points, jsdom constraints, and pre-commit
checklist so future sessions can add rules end-to-end without
re-investigating the wiring.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Rename /craft skill to /shape and make it planning-only (no code output)
- Add /impeccable craft sub-command that calls /shape, loads references,
builds with visual iteration until delightful
- Replace all " -- " (em dash substitutes) with proper punctuation across
all skill files and index.html
- Move v1.6 changelog entry to "View older releases" section
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The CLI and detection engine now live in pbakaus/impeccable-detect
(published as 'impeccable' on npm, BSL-1.1). This repo is purely
Apache 2.0: skills, prompts, website, and build system.
- Remove bin/ (CLI moved to CLI repo)
- Remove README.npm.md (moved to CLI repo)
- Remove @impeccable/detect dependency, add impeccable dependency
- Set package.json to private (no longer published to npm)
- Update all references from @impeccable/detect to impeccable
- Update CLAUDE.md, NOTICE.md, FAQ, and changelog
- Rebuild all provider skill distributions
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
