llm-cost-optimizer
🔌Pluginaneeba-pixel/claude-skills
|Cut LLM API spend via model routing, prompt caching, prompt compression, and per-feature cost observability. Use when AI costs are too high, choosing between models, or launching an AI feature without cost architecture. NOT for RAG design or prompt quality (separate skills).
Part of
aneeba-pixel/claude-skills
Installation
/plugin marketplace add alirezarezvani/claude-skills/plugin install llm-cost-optimizer@claude-code-skillsMore from this repository10
Snowflake SQL, data pipelines (Dynamic Tables, Streams+Tasks), Cortex AI functions, Snowpark Python, and dbt integration. Includes query helper script, reference guides, and troubleshooting.
Marketplace
Hypothesis testing, A/B experiment analysis, sample size calculation, and confidence intervals. 3 stdlib-only Python tools: Z-test/t-test/chi-square with effect sizes, sample size calculator with power tradeoffs, and Wilson score confidence intervals.
Google Workspace administration via the gws CLI. Install, authenticate, and automate Gmail, Drive, Sheets, Calendar, Docs, Chat, and Tasks. 5 Python tools, 3 reference guides, 43 built-in recipes, 10 persona bundles.
Reverse-engineer any codebase into a complete PRD. Frontend (React, Vue, Angular, Next.js), backend (NestJS, Django, Express, FastAPI), and fullstack. 2 Python scripts (codebase_analyzer, prd_scaffolder), 2 reference guides, /code-to-prd slash command.
Chief Data Officer advisory for startups: AI training data audit (origin × class × use-case matrix with GDPR Art. 6 + EU AI Act citations), data product strategy picker (warehouse vs lakehouse vs mesh + 6-layer build-vs-buy + 12-month sequencing), data asset valuator (strategic value 0-10 + M&A multiplier with carve-out penalties + 3 ranked productization paths). 4 references answering one decision each: training rights, data product strategy, customer-data-as-asset, data team org evolution. Standalone-installable; also bundled in c-level-skills. Strategic only — does not duplicate engineering data skills.
Research orchestrator (hybrid router + fallback). Deterministic SIGNALS classification routes to 6 specialists (pulse/litreview/grants/dossier/patent/syllabus) at >=2 signals, else runs own 8-step plan-decompose-search-synthesize-cite fallback. Routing transparency mandatory. Path-B from megaprompt 13.
Ultra-compressed communication mode. Cuts token usage 20-50% (75% upper bound) by dropping filler, articles, pleasantries, and hedging while keeping full technical accuracy. Derived from Matt Pocock's MIT-licensed caveman with: (1) 3 stdlib Python tools (deterministic compressor, token-savings estimator with $/Mtok cost extrapolation, lint that detects banned vocab with code-block + exception-zone whitelisting), (2) 3 references citing 7-8 sources (compression principles, when caveman backfires, companion tooling), (3) cs-caveman-mode persona agent + /cs:caveman slash command. Matt's persistence rules + auto-clarity exception preserved verbatim per MIT.
End-to-end chaos engineering discipline: design experiments with hypothesis + steady-state metric + blast radius + abort criteria, calculate risk score against error budget, and generate blameless postmortems. 3 stdlib Python tools (experiment_designer, blast_radius_calculator, experiment_postmortem), 4 references on chaos principles + experiment design + 7-attack taxonomy + tooling landscape (Chaos Toolkit/Mesh/Litmus/Gremlin/AWS FIS/DIY), templates, and /chaos-experiment slash command. Composes with feature-flags-architect (kill switches as abort triggers) and kubernetes-operator (chaos targets).
Patent prior-art + IP landscape skill. FTO/novelty/family-resolver via 3-pass Jaccard heuristic. Research-pack convention. Path-B from megaprompt 12.