elchananteitzwork-a11y
Community@elchananteitzwork-a11y
Governance skills enforcing feature-flag cutovers, provider integration contracts, R&D documentation, confidence-tiered claims, and Anthropic cost controls in a supplement-intelligence codebase.
Agent Skills by elchananteitzwork-a11y
Showing 6 vetted skills indexed across 1 GitHub repositories.
elchananteitzwork-a11yfeature-flag-staged-cutover
Enforces recorded flag, rollback, and validation evidence before milestone cutover.
elchananteitzwork-a11ydata-provider-integration-contract
Enforces shared caching, error, confidence, and provenance patterns for external data-provider integrations.
elchananteitzwork-a11yrd-document-generator
Generates milestone Research and Design documents following a fixed seven-section template.
elchananteitzwork-a11yconfidence-tiered-extraction
Enforces confidence-tier labeling and bans revenue, ROI, and success-probability claims in user-facing output.
elchananteitzwork-a11yllm-cost-rate-governance
Validates Anthropic SDK call sites for rate-limit handling, retries, and graceful degradation.
elchananteitzwork-a11yproduct-hunt
Runs a 12-agent pipeline to score and rank Amazon product opportunities for a niche.
Frequently Asked Questions About elchananteitzwork-a11y
FAQPage SchemaWhat tasks can I enforce using elchananteitzwork-a11y's skills?▼
You can gate feature-flag cutovers on recorded rollback plans and validation evidence, standardize external data-provider integrations on shared cache and error patterns, generate milestone Research & Design documents, enforce CONFIRMED/INDICATED/LIMITED confidence tiers on user-facing claims, and require retry and degradation policies on Anthropic SDK calls.
Who should use these skills?▼
Planners opening milestones, implementing engineers wiring feature toggles or provider clients, and anyone editing Anthropic prompts, extraction pipelines, or verdict-formatting code in the supplement-intelligence repository. They target code paths under lib/ engines such as science-engine, regulatory-engine, thesis-engine, and ai-interpretation.
When should each skill be triggered or skipped?▼
Trigger feature-flag-staged-cutover before marking milestones live, data-provider-integration-contract when calling providers like Keepa or ClinicalTrials.gov, and llm-cost-rate-governance when importing @anthropic-ai/sdk. Skip them for routine bug fixes, UI-only changes, internal scoring math, and tests that merely mock providers.
What content restrictions does confidence-tiered-extraction impose?▼
It forbids revenue predictions, ROI claims, and probability-of-success language anywhere in generated output. Every user-facing claim, score, or verdict shown to a founder must carry an explicit CONFIRMED, INDICATED, or LIMITED confidence tier, applying to prompts, extraction pipelines, and interpretation-formatting code.
What are the prerequisites and dependencies for these skills?▼
They assume the repository's existing architecture: provider-cache and provider-errors modules, engine directories under lib/ and supplement-intelligence/lib/, and the Anthropic SDK for LLM calls. The skills declare disallowed-tools: Agent, meaning they run without delegating to an Agent tool during enforcement.