* fix(37signals-way): replace the dead Shape Up Amazon link
ASIN B09ZSY1MWP returns Amazon's "Page Not Found". Shape Up has no
English Amazon listing at all, only a Portuguese Kindle edition, so
there is nothing to swap the ASIN for. The book is free at
basecamp.com/shapeup, which is now where the link points.
Updated in the skill's Further Reading, both generated plugin copies
(re-ran scripts/generate-plugins.sh, which reproduces the edit exactly),
and the two README mentions. The landing page carries the same fix.
The four other 37signals books in that section all resolve to the right
title and author.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012todV6BsX7FQor88FtgbAa
* fix(links): repair four more dead Amazon links found by a full sweep
Checked all 98 unique Amazon ASINs across the repo against the live
product pages. Five were dead; Shape Up was the first. The other four:
- crossing-the-chasm — "Inside the Tornado" pointed at 0887307760, dead.
Now 0060745819, the live Collins Business Essentials edition.
- obviously-awesome — April Dunford's "Sales Pitch" pointed at
1999023048, dead. Now 1999023021, the paperback.
- refactoring-ui — "Refactoring UI" pointed at B0BLJ7MC21, dead. The book
has never had an Amazon listing; it is sold at refactoringui.com, which
is now the link.
- ios-hig-design — "Designed by Apple in California" pointed at
1942303118, dead. It is out of print, Apple's own page redirects to the
homepage, and the only remaining Amazon listing is a Scandinavian
marketplace copy. The title keeps its entry and loses its link.
Plugin copies regenerated with scripts/generate-plugins.sh. The other 93
ASINs all resolve to the right title and author.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012todV6BsX7FQor88FtgbAa
* chore(skills): bump patch versions for the link corrections
CLAUDE.md requires a version bump whenever a SKILL.md changes; a small
fix takes a PATCH increment. All five skills touched by the dead-link
repairs get one, and the generated plugin copies were refreshed.
- 37signals-way 1.3.0 -> 1.3.1
- crossing-the-chasm 1.4.0 -> 1.4.1
- obviously-awesome 1.4.0 -> 1.4.1
- refactoring-ui 1.5.0 -> 1.5.1
- ios-hig-design 1.5.0 -> 1.5.1
Marketplace versions are left alone: they are auto-synced to the latest
GitHub release by scripts/sync-marketplace-versions.sh, not hand-edited.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012todV6BsX7FQor88FtgbAa
---------
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
New book-style skill (65th) for auditing, writing, and improving developer
documentation, based on Google's Developer Documentation Style Guide and the
Technical Writing One/Two courses, plus Keep a Changelog for release notes.
SKILL.md (231 lines) covers reader/doc-type fit, voice, sentences and words,
structure, procedures and code, API reference, changelogs, and a three-mode
protocol (audit / write / improve) with a fixed audit report format. Seven
references (1.8k-2.7k prose words) carry the detail; audit-checklist.md holds
the frozen 60-rule table (R/V/W/S/P/A/N) whose IDs every finding cites.
Notable design points:
- First non-book source in the repo: "## About the Source" replaces "About the
Author", with CC BY 4.0 attribution; CLAUDE.md documents the variant.
- Two rule layers: structural rules apply to any language, [EN]-tagged rules
are skipped for non-English docs (diagnostic row 10 auto-passes).
- Facts outrank style: every command, flag, and parameter must be traced to
code or the user; unverifiable facts become TODO(verify) rather than
guesses or deletions.
- Anything not on a Google page is labelled (inferred) or (convention).
Validated with the skill-creator eval loop (3 tasks x with/without skill,
Opus subagents, mechanical grader): with-skill 100% of assertions vs 42%
baseline in iteration 2. Registered in code-craftsmanship; counts, README,
CLAUDE.md, and EXAMPLES.md updated; generated trees regenerated.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Records the rule in the repo's guidance file: tests that restate the
implementation prove nothing, break on refactor, and should be deleted
rather than repaired. Notes the characterization-test carve-out so the
working-with-legacy-code examples aren't caught by it.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Both journeys were executed end to end in scratch projects with every
constituent skill treated as uninstalled, to test the standalone path.
Highest-severity defect, hit by both: references/artifact-templates.md
shipped skeletons for only 2 of the 5-7 docs/ files each journey writes,
while Operating Rule 7 orders creating missing files "from the full
skeleton". The remaining headings live only in docs/ARTIFACT-REGISTRY.md,
which is maintainer-facing and ships to nobody — so a standalone run
invented them and silently dropped canonical sections. Both templates now
carry every skeleton they write.
conversion-optimization:
- §1.4 leak-value formula was dimensionally wrong: raw lost users x LTV
always ranks top-of-funnel first, defeating its own instruction. Now
gap-to-benchmark x downstream rate x value, i.e. recoverable money.
- §7 mandated a guardrail but sized only the primary metric; at funnel
volumes the guardrail often cannot reach power. Added a sizing step
with the monitored-not-gating fallback and a fast proxy.
- Phase 1's GATE was unsatisfiable on a partly instrumented funnel;
now gates on localizing the leak, with dark sub-steps awaiting-evidence.
- Homes added for outputs that had none: Brand Script, Price Metric, and
a Proposed message/CTA column so rewrites stop overwriting the
before-state that Phase 7 measures against.
- Phase 6 no longer demands checkout semantics in payment-free flows.
architecture-optimization:
- Rule 8 read literally auto-reverted every Phase 7-8 change, since
timeouts, breakers and CI gates move no latency baseline; now scoped to
changes aimed at a baseline.
- Phase 1 fused two independent nets behind one gate; split so pinned
behavior unlocks 2-4 and a baseline unlocks 5-6.
- Phase 6 named an input Phase 5 never produces; now cites the Phase 1
profile and says to enable pg_stat_statements/auto_explain first.
- Phases 5 and 6 may swap when the profile shows the database dominating
— caching ahead of an N+1 fix is the skill's own named mistake.
- Phase 1 now creates TECH-DEBT.md, which it was already told to write to.
- methods.md gains pool sizing via Little's law, cache-stampede guidance,
EXPLAIN (ANALYZE, BUFFERS), and composite index column order.
Also: measured load supersedes the QPS estimate; benchmarks added for the
funnel steps that had none; scope qualifiers so "every list endpoint" and
"every outbound call" mean the in-scope flows; tracker records fallback
vs skills-installed mode and what awaiting-evidence is waiting for;
optional phases state their standalone method instead of implying one.
create-website now routes to conversion-optimization (the only sibling
border that was one-way).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
marketplace.json: both journeys added to the metaskills collection, with
counts, description, and keywords updated. Version fields untouched —
they are owned by the release workflow.
CLAUDE.md, README.md, EXAMPLES.md: counts 62 -> 64 and 12 -> 14
throughout, catalog table rows, install lists, two Skill Details
sections, and two Guided Journeys scenarios (94 scenarios, matching the
README claim). CLAUDE.md also notes that the two optimization metaskills
ship references/methods.md for standalone operation.
Regenerated the IDE mirrors and plugin trees via scripts/sync-ide-skills.sh.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Sentence 4 of each metaskill description routes to siblings as negative
triggers; the two new journeys create six real borders. Descriptions are
capped at 1024 chars, so each edit trims existing text to make room.
improve-website, grow-website, improve-app -> conversion-optimization
improve-code-quality, remove-technical-debt, design-code-architecture
-> architecture-optimization
PATCH bump on each.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two new guided journeys, both designed to run standalone — every phase
carries its full method inline (Brief + references/methods.md), so the
journey works with no constituent skill installed.
conversion-optimization (7 phases): one leaking conversion flow —
landing, signup, checkout, or in-app onboarding. Measure the leak
(lean-analytics), research why (cro-methodology), then fix message,
offer, proof, and friction in that order, and prove each fix with a
pre-committed test. Draws on LIFT, the MECLABS heuristic, Fogg's B=MAP,
Baymard checkout findings, and Kohavi on trustworthy experiments.
architecture-optimization (8 phases): a working codebase grown slow and
tangled. Measure before optimizing, pin before restructuring — structure
phases first so the optimizations are small diffs, then the measured
bottleneck and the data layer. Draws on Gregg's USE method, Amdahl's and
Little's laws, and performance budgets as CI gates.
Registry: adds docs/FUNNEL.md and docs/PERFORMANCE.md as new artifacts
and records the two journeys against the artifacts they extend.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Package the 10 collections as dual-format plugins and add an all-in-one
bundle, all generated from .claude-plugin/marketplace.json:
- plugins/<collection>/plugin.json — Agent Plugins v1.0.0 manifest next to
the existing .codex-plugin/plugin.json (Codex output byte-identical)
- plugins/<collection>/skills/ — real copies instead of symlinks: the spec
requires every package path to resolve inside the plugin root, so
conforming clients would skip out-of-root symlinked skills
- plugins/wondelai-skills/ — all-in-one Agent Plugin with all 62 skills
(Agent Plugins only; not offered in the Codex marketplace)
- rename scripts/generate-codex-plugins.sh -> generate-plugins.sh; validate
zero symlinks + SKILL.md presence + manifest shape
- CI change detection now catches untracked generated files
(git status --porcelain instead of git diff --quiet)
- README: new 'Via Agent Plugins' install section; CLAUDE.md: updated
generated-tree docs + staleness note (copies need re-sync after skill edits)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GhHsvxpceWjX3dozmkhYbJ
Phase Invoke lines, Operating Rule 4, the optional-phases footer, and
forward-routing lines now say "use the <slug> skill" / "load the
phase's skill" instead of a bare slug, so executing agents reliably
trigger the constituent skills. PATCH version bump on all 12 metaskills.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GhHsvxpceWjX3dozmkhYbJ
The workflow ran generate-codex-plugins.sh via sync-marketplace-versions.sh
but only staged .claude-plugin/marketplace.json, leaving plugins/ and
.agents/plugins/ stale after each release (they sat at 1.4.0 through
v1.5.0 and v1.5.1).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_011D9pvKw5jEaufBm72FPAH4
improve-app v1 mixed seven code-quality phases with UX work, making it a
near-duplicate of improve-code-quality. v2 is the product-experience
journey: jobs-to-be-done (GATE) → ux-heuristics → design-everyday-things →
refactoring-ui → microinteractions → made-to-stick (in-app copy) →
influence-psychology (honest in-app persuasion) → high-perf-browser
(perceived speed) → steve-jobs-design-review. All code, test, and
production-hardening concerns now route to improve-code-quality /
remove-technical-debt.
- Artifacts shift from TECH-DEBT/TESTING/ARCHITECTURE/RELIABILITY to
CUSTOMER/DESIGN/POSITIONING/PRODUCT/EXPERIMENTS; registry lists updated
- New rule-8 guardrail: evidence before redesign; persuasion stays honest
- grow-app 1.0.1: drop the stale claim that improve-app fixes reliability
- README table/details and EXAMPLES entry rewritten for the new scope
- Both skills pass quick_validate; Operating Rules 1-7 still verbatim-
identical across the 12
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JSE3k8WcAFcrbzWbzt6rnv
New skill archetype: interactive orchestrators that drive a whole journey
phase by phase — invoking the constituent skills, asking the user at every
decision point, and recording results as UPPERCASE artifacts in the user's
project docs/ folder (shared across journeys, plus a per-journey tracker
for cross-session resume).
The 12 journeys (one per skills.wondel.ai guide): create/improve/grow ×
business/website/app, improve-code-quality (vibe-coded prototype →
production), remove-technical-debt, design-code-architecture.
- Each SKILL.md: Journey Map, verbatim-shared Operating Rules (resume-first,
intake, phase entry/exit, skill-invocation fallback, artifact discipline,
journey guardrails), six-field phases with checkable done-conditions,
optional phases, sibling-routing descriptions with negative triggers
- references/artifact-templates.md per skill + canonical registry at
docs/ARTIFACT-REGISTRY.md
- New 10th marketplace collection `metaskills`; counts 50 → 62; CLAUDE.md
documents the metaskill format; README catalog/details/install; EXAMPLES
Guided Journeys section; IDE mirrors + Codex plugins regenerated
- All 12 pass skill-creator quick_validate; smoke-tested end-to-end
(create-business and remove-technical-debt: intake, GATE, commit
discipline, awaiting-evidence pause, resume-first)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JSE3k8WcAFcrbzWbzt6rnv
Add a call-to-action section linking to
https://skills.wondel.ai/work-with-us/, placed between the Skills
Ecosystem section and the Copyright & Disclaimer footer. Describes the
three service tiers (skill sprints, custom builds, skills library)
without hardcoding prices, keeping the linked page as the source of truth.
Claude-Session: https://claude.ai/code/session_01KWKbscRFaitbTd912vV5VE
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Tagline now: "Claude, Claude Code, Claude Cowork, Codex, Cursor, OpenClaw,
Hermes Agent and other agentskills.io-compatible agents" across README,
CLAUDE.md, .cursorrules, and copilot-instructions. Install headings and
commands referencing "Claude Code" are untouched.
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Update the "agent skills for Claude Code and agentskills.io-compatible
agents" tagline to name Codex, OpenClaw, and Hermes Agent explicitly,
across README, CLAUDE.md, .cursorrules, and copilot-instructions.
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Codex adopted the open SKILL.md standard, so the skills already work there;
this adds the Codex *plugin* standard for parity with the Claude marketplace,
generated from .claude-plugin/marketplace.json as the single source of truth.
- scripts/generate-codex-plugins.sh: emits plugins/<collection>/.codex-plugin/
plugin.json + skills/ symlinks + .agents/plugins/marketplace.json (9
collections, 51 skill links), reusing each collection's Claude metadata.
- sync-ide-skills.sh: add .agents/skills flat mirror (5th target) + call the generator.
- sync-marketplace-versions.sh: regenerate Codex plugins so versions track the
release (all at 1.4.0).
- Docs: README Codex install section; CLAUDE.md structure + add-skill step 6.
All generated (never hand-edited); versions auto-track releases via the sync workflow.
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Marketplace plugin versions lagged behind the skill content, so plugin
installs were not offered the v1.4.0 refresh. Align them to the GitHub
release version and keep them synced automatically.
- Set marketplace.json metadata.version + all 9 plugin collections to
1.4.0 (matches release v1.4.0). Collections bumped from 1.1.0/1.2.0;
top-level realigned from an ad-hoc 1.6.0 to track the release tag.
- Add scripts/sync-marketplace-versions.sh (sets every version to a given
release/tag; idempotent).
- Add .github/workflows/sync-marketplace-version.yml: on each published
release (and manual dispatch), run the script with the release tag and
commit the result to main.
- Document the automation in CLAUDE.md (versions are no longer hand-edited).
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* refactor(skills): lift all 50 skills to 9-10 vs writing-great-skills rubric
Apply the writing-great-skills rubric to every skill via a shared fix
playbook: cut no-op platitudes, enforce single-source-of-truth, add
when-to-read pointers, de-synonym descriptions, add concrete scoring
bands, and trim sprawl -- without removing framework substance.
- Net -751 lines (857 ins / 1608 del) across 50 SKILL.md + 5 references
- Removed 48 duplicative "Reference Files" rosters (kept inline pointers)
- Cut ~67 net platitude "Ethical boundary" lines + 15 "Why it works" no-ops
- Fixed influence-psychology 6-vs-7 principle bug (Unity now triggerable)
- Bumped each skill version (MINOR); README/marketplace/CLAUDE untouched
- All 50 YAML frontmatters validated; independent re-score 9.0-9.4
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
* refactor(skills): finish Reference Files removal + document new format
Reconcile the writing-great-skills style as canonical (per maintainer):
- Remove the 2 remaining "## Reference Files" rosters (clean-code,
team-topologies); every reference stays reachable via inline
when-to-read pointers (verified 0 orphans across all 50).
- Update CLAUDE.md's documented SKILL.md format: drop the "## Reference
Files" section, document inline when-to-read pointers, mark "Why it
works"/"Ethical boundary" as keep-only-if-behavior-changing, and
Scoring as concrete bands tied to the diagnostic.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
* fix(skills): address Codex review on PR #11 (scoring math + INP semantics)
- influence-psychology: Scoring counted 7 diagnostic rows "one point each"
but the bands required 9-10 — unreachable (capped at 7). Drop the
row-counting; score against qualitative bands (added a 7-8 band for a
full reproducible scale) using the diagnostic as the checklist.
- high-perf-browser: INP was described as blocking "the first input"
(FID semantics). INP covers all interactions across the page lifecycle
(matches references/core-web-vitals.md). Fixed the Why-it-works line
and the Key-insights bullet.
- Both bumped PATCH 1.4.0 -> 1.4.1.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
* fix(skills): repair unreachable scoring bands across 6 skills
Codex flagged that several Scoring sections counted N diagnostic rows at
"1 point each" while the bands required 9-10 — making the top band
mathematically unreachable (max < 10). Root cause was a buggy example
("1 point per satisfied row") in CLAUDE.md's documented format.
- system-design, microinteractions, refactoring-patterns, hooked-ux:
score = round(passed / N × 10) so full marks reach 10.
- blue-ocean-strategy, traction-eos: map row-count to the existing bands
(which already span 0-10) instead of "1 point each".
- CLAUDE.md: fix the Scoring guidance to require a 0-10-spanning rule and
warn against "1 point per row" when there are fewer than 10 rows.
- PATCH-bumped the 6 skills. (influence-psychology, same bug class,
was fixed in ed83619.)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019ygbNRWuZjaizDpcLEQ6jj
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(skills): optimize all 50 descriptions for auto-activation
Add plain-language, problem-framed trigger phrases to every skill's
frontmatter `description` (the auto-activation surface) so skills fire
when users describe their problem naturally, without naming the skill or
its jargon. Scored each against a skill-creator-derived 0-10 rubric
(avg 7.7 -> 10/10); the consistent gap was layperson phrasings.
- 50 SKILL.md: +3-5 plain-language triggers each, format preserved,
all <=1024 chars, cross-refs valid, MINOR version bumps
- README.md: one problem-framed "Use when you need to:" bullet per skill
- marketplace.json: +42 keywords across 9 plugins, MINOR version bumps
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014yfjVfs1kC4UKFd5cjVnuw
* fix(skills): address Codex review on system-design and 37signals-way
- system-design: scope bare "interview prep" trigger to
"system design interview prep" to avoid matching unrelated
(behavioral/product/sales) interview-prep requests
- 37signals-way: restore "fat marker sketch", "underdo the competition",
and "opinionated software" triggers dropped in the prior rewrite;
they remain core concepts, so keep them alongside the new layperson phrases
No version bump: refines the same unmerged description change.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014yfjVfs1kC4UKFd5cjVnuw
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Rated every SKILL.md against a 10-point rubric derived from the
official skill-creator skill (description quality, progressive
disclosure, lean body, imperative voice with reasoning, reference
hygiene, actionability, template consistency) and applied a uniform
"meaningful tightening" pass: author bios cut to 2-3 sentences,
table cells to fragments, duplicate See: links collapsed, stub
sections removed, restating bullets pruned.
Result: 126,603 -> 97,780 words (-22%) and 13,222 -> 10,618 lines
(-19%) across all 42 SKILL.md files with zero framework sections,
scoring rubrics, diagnostics, or quoted copy patterns lost. All
files now <=394 lines (was <=472). Also fixes a factual inversion
in blue-ocean-strategy (Novo Nordisk case contradicted its own
reference file) and a 30-50%% typo in scorecard-marketing's
description. Every skill version bumped MINOR.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
New UX/Design skill distilling Steve Jobs' documented review practices
from Isaacson's biography, Segall's "Insanely Simple", and Kocienda's
"Creative Selection": simplicity audits, the no list, design-is-how-it-
works, end-to-end experience ownership, demo culture, and a review
protocol with binary verdicts (insanely great / not done).
Includes 5 reference files (review protocol, simplicity & focus,
end-to-end experience, demo culture, case studies) and IDE symlinks
for all 4 editor integrations.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adds missing symlinks across .claude, .cursor, .windsurf, and .pi for
skills that hadn't been synced yet (37signals-way and 20 others).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New product development skill synthesizing 37signals philosophy across
three books. Covers shaping, betting, six-week cycles, small teams,
build-less philosophy, and opinionated software with 7 reference files
including dedicated UX/UI/copy guide and case studies.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Audited all 41 skills against Anthropic's skill-creator SKILL.md requirements.
Two gaps found and fixed:
- Added Table of Contents to 88 reference files >300 lines (was 10/98)
- Expanded all 41 descriptions from avg 51 to avg 76 words with more trigger
phrases and "Also trigger when..." clauses for better skill activation
- Bumped MINOR version on all 41 skills
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Custom cursors now require explicit user confirmation before implementation.
This prevents agents from adding custom cursors by default, which can hurt
usability when misapplied.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Link all 41 skill names in the Available Skills table and Skill Details
headings to their pages on skills.wondel.ai, and add landing page link
to the opening line.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
domain-driven-design: Replace `com.app` package placeholder (flagged as
malicious URL by Agent Trust Hub) with `myapp` prefix.
release-it: Add safety notes and guardrails to chaos engineering section.
Replace raw shell commands with references to dedicated chaos tooling
(Gremlin, Litmus, AWS FIS). Clarify that failure injection is a planning
concept, not an action for autonomous execution.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
New persona sections: Software Engineers (7 examples), Tech Leads & Architects (7 examples).
Added examples to Startup Founders (+2), Product Managers (+4), UX/UI Designers (+2).
Updated Skill Index with all 15 new skills.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
New UX/Founder skills: mom-test, inspired-product, lean-ux, continuous-discovery, microinteractions
New Code Quality skills: clean-code, refactoring-patterns, software-design-philosophy, pragmatic-programmer, domain-driven-design
New Systems/Architecture skills: ddia-systems, system-design, clean-architecture, release-it, high-perf-browser
Each skill includes SKILL.md + 6-7 reference files. Added two new marketplace collections (code-craftsmanship, systems-architecture) and updated existing collections.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Compress Nielsen's 10 heuristics from detailed per-heuristic sections
(tables, copy patterns, ethical boundaries) to concise summaries.
Full details remain in references/nielsen-heuristics.md.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
New behavior design skill with SKILL.md and 7 reference files covering
the Fogg Behavior Model, Ability Chain, prompt design, Tiny Habits method,
motivation waves, product applications, and case studies. Updates
marketplace.json, CLAUDE.md, and README.md to reflect 26 total skills.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add $schema, top-level description, and per-plugin metadata (version,
author, homepage, repository, license, category, keywords) to match
the official anthropics/claude-plugins-official format. Bump version
to 1.2.0. Passes `claude plugin validate .` cleanly.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Michal Jaskolski