The library

spec-miner · Creating specifications for legacy systems, mapping dependencies, or documenting existing code without docs

better-all · Running multiple async operations with complex interdependencies using automatic DAG optimization

spring-boot-engineer · REST APIs, Spring Security 6, reactive WebFlux, Spring Data JPA, or microservices architecture

sql-pro · Query optimization, indexing strategy, performance tuning, or advanced SQL patterns

simple-fetch · Reliable HTTP requests with automatic retries, error handling, or timeout management

bio-annotation · Gene annotation, sequence analysis, or biological data enrichment pipelines

swift-expert · Swift concurrency, SwiftUI, async/await, actor isolation, or Apple platform development

terraform-engineer · Managing multi-environment infrastructure as code where state consistency, dependency ordering, and secrets isolation are critical.

manim-skill · Iterating on complex animations by rendering frames and diagnosing visual issues before committing to final video production.

test-master · Designing comprehensive test architectures across unit, integration, E2E, performance, and security when coverage gaps risk production failures.

env-handler · Securing API keys and secrets in .env.local while distributing safe templates in .env across the team.

typescript-pro · Building type-safe APIs where branded types and discriminated unions catch misuse at compile time rather than runtime.

vue-expert-js · Building Vue 3 Composition API components in JavaScript with JSDoc types when TypeScript setup adds overhead.

vue-expert · Building Vue 3 Composition API components with strict type safety across composables, routing, and state management.

websocket-engineer · Building real-time bidirectional communication where HTTP polling would waste resources and introduce latency.

quality-assurance · Designing QA workflows that balance automated gate checks with manual exploration when shipping must meet quality bars.

wordpress-pro · Extending WordPress with custom post types and REST endpoints when off-the-shelf plugins don't fit the domain model.

implementing-llms-litgpt · Production-grade fine-tuning without abstraction layer overhead; ideal for understanding model internals.

quality-reviewing · Catching ecosystem mismatches before they reach production; complements automatic hook with external verification.

mamba-architecture · Long sequences (100K+ tokens), streaming inference, or memory-constrained deployments beating Transformer quadratic scaling.

nanogpt · Learning transformer internals, quick prototyping, or experimenting with variants without framework overhead.

rwkv-architecture · Combining RNN efficiency with Transformer quality when linear-time inference matters.

distributed-llm-pretraining-torchtitan · Large-scale pretraining where single-node limits are exceeded and distributed coordination is unavoidable.

huggingface-tokenizers · Fast, parallel tokenization when model-specific vocabularies and special tokens matter.

sentencepiece · Multilingual models or custom vocabularies where language-agnostic approach beats language-specific tokenizers.

ecosystem-alignment · Keeping a custom Claude Code setup in sync with upstream platform updates, preventing drift and discovering new capabilities to adopt.

peft-fine-tuning · Fine-tuning large models (70B+) on consumer hardware by training only 0.17% of parameters in a 6MB adapter, enabling cost-effective task-specific customization without full-model training.

rca-distributed-systems · Production incidents in microservice meshes (Kubernetes, Istio, Linkerd) where the fault could originate in any of 10-100 services and the causal path runs through network hops, retries, and timeouts.

nnsight-remote-interpretability · Running the same interpretability code on GPT-2 locally and Llama-405B remotely without code changes, enabling scalable mechanistic interpretability research on massive models.

pyvene-interventions · guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention framework

sparse-autoencoder-training · guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable f...

push · Solving push challenges

transformer-lens-interpretability · guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints a...

nemo-curator · GPU-accelerated data curation for LLM training

ray-data · Batch inference and preprocessing on 100GB+ datasets across multi-node clusters.

riverpod-cancel · Flutter apps cancelling requests when user navigates or triggers rapid refreshes.

grpo-rl-training · Teaching specific output formats (XML, JSON) and verifiable tasks without preference pairs.

miles-rl-training · Training 1TB+ MoE models with speculative RL for 25%+ rollout speedup.

openrlhf-training · Scaling PPO/GRPO/RLOO/DPO training to 70B+ models with multi-node vLLM.

model_engineer · Designing end-to-end ML systems from task spec to production deployment.

simpo-training · Quick preference optimization without reward model or RL infrastructure.

claude-typescript-sdk · Integrating Claude into TypeScript backends and Node.js scripts.

slime-rl-training · Research-grade RL training with flexible reward functions and algorithm variants.

sentry-known-pitfalls · Quick diagnosis of Sentry setup failures and error categorization issues.

research-mission-generator · Decomposing business goals into actionable research programs.

torchforge-rl-training · PyTorch-native RL training with hardware acceleration and custom loss functions.

phpstan-resolver · Migrating PHP codebases to strict type checking incrementally.

fine-tuning-with-trl · Multi-phase RLHF pipelines (SFT→Reward→PPO) where you control each alignment stage.

verl-rl-training · Production math/reasoning tasks (GSM8K, MATH) where you need proven RL algorithms at scale.

pnpm · Monorepos needing deterministic builds, strict dependency isolation, and centralized version catalogs.