The library

confusion-matrix-generator · Verification workflows where you need both supporting evidence and known conflicts for a claim, e.g. deal history, research synthesis, coaching notes.

add-component · Maintaining consistent Vue component structure across a design system.

network-config-validation · Pre-flight validation of generated or hand-edited network config before change windows.

nextjs-skills · Onboarding Claude to consistent Next.js conventions without manual documentation.

uninstall-extension · Safe extension cleanup that retains user data and avoids manual file deletion.

diff-feature-branch · Quick review of changes before merge or code-review.

claude-skill-lord · Managing distributed Claude skills across projects.

claude-skill-lint · Production skill repos need deterministic validation before publishing; catches hallucinated context bugs CI cannot see.

analyst · Killing bad ideas early; composite scoring prevents false positives on cool-sounding unfeasible proposals.

nodejs-keccak256 · Ethereum/EVM teams; one wrong hash breaks selectors/signatures/storage—non-obvious until mainnet.

cto-advisor · CTOs need fast principled evaluation of tech choices; beats design committee meetings.

obsidian-markdown · Creating notes with wikilinks, embeds, and callouts that render correctly in Obsidian vaults.

onboarding-cro · Shortening time-to-value by removing friction before the aha moment and measuring activation drop-off step-by-step.

skill-installer · Installing Stashwise-generated Claude skills with cryptographic signature verification and safe unpacking.

database-schema-designer · Designing normalized relational schemas with ERD diagrams and DDL generation from high-level requirements.

qa · Running deterministic test suites and coverage checks to prevent regressions before code merges.

resolver · Debugging production issues by tracing logs, reproducing errors, and identifying root causes systematically.

web-design-guidelines · Ensuring UI consistency, accessibility (WCAG AA), and performance across responsive designs.

decide · Prevents regressions by preserving dissent from rejected options that often prove correct at kill-criteria time.

reviewer · Distinguishes blocking issues from nits—allows teams to merge without blocking on style when logic is solid.

security · Severity classification separates critical auth issues from style nits—allows teams to merge without blocking on low-severity findings.

vp · Sits above individual cells—resolves conflicts without forcing cells to restructure or hide constraints.

nansen-web-searcher · Provides on-chain analytics that free block explorers cannot—holder distribution, interaction patterns, risk signals.

anti-deception · Catches narrative contradictions invisible to humans reading sequentially—flags shifts in story that seem plausible alone.

code · Code generation when you need to catch behavioral changes hidden by passing tests.

memory · Cross-turn pattern detection when distinguishing real signals from projection.

perl-patterns · Perl projects when modernizing pre-5.36 code to idiomatic v5.36+ standards.

reasoning · Complex analysis when you need to avoid causal shortcuts and premature conclusions.

docs-maintenance · Keeping project documentation fresh after feature completion to reduce AI context usage.

execute · Converting an approved decision into weekly-checkable milestones with named owners and kill criteria.

debugging-with-tools · Investigation-heavy debugging when tool-driven evidence (search/logs/diffs) beats guesswork.

executive-mentor · Founders/execs preparing for board meetings or making irreversible decisions needing honest pre-mortems.

refactor-pass · Cleaning up code after feature work by removing dead code and straightening logic.

prediction-market-oracle-research · When you want to use market-implied probabilities to inform strategic decisions while controlling for liquidity, resolution rules, and manipulation.

fda-qsr-audit-prep · When you need to audit medical device QSR posture before internal review, FDA inspection, or Form 483 response.

prediction-market-risk-review · When a trading agent, portfolio tool, or oracle integrations touches venue auth, portfolio data, or execution capabilities.

financial-analyst · When you need to interpret P&L, cash flow, balance sheet trends and build financial models.

financial-health · When you need a snapshot of cash position, growth efficiency, and runway to understand financial sustainability.

product-marketing-context · When launching a new marketing effort and you need to capture positioning, ICP, problems, and differentiation once so every marketing skill reuses the same foundation.

weather · When you need quick weather for a location and want zero API keys or rate limits.

github-pr · When you need to create PRs programmatically and want consistent metadata (title, description, reviewers) across many PRs.

founder-mode · Founders who want one-command routing instead of picking from 23 slash commands.

prompt-optimizer · Refining prompts before execution when you want routing and component matching advice.

freeze · Founders who want to prevent second-guessing after irreversible or high-cost decisions.

full-page-screenshot · Capturing visual proof of features, UI changes, or cross-platform rendering without manual cropping.

python-backend-orchestrator · Coordinating parallel async work across multiple Python services in a background job.

gc-review · Quick legal risk screening before executive sign-off or escalation.

python-testing · Building regression test suites that catch regressions before commit via deterministic TDD cycles

writing-for-llms · Prompt engineering that avoids anti-pattern traps by showing only correct patterns and front-loading critical info

generate · Generating on-brand content at scale by enforcing guidelines and voice rules at generation time