性能
性能分析与优化
prompt-engineering-expert
TomsTools11
Advanced expert in prompt engineering, custom instructions design, and prompt optimization for AI agents
huggingface-tokenizers
eyadsibai
Use when "tokenizers", "HuggingFace tokenizer", "BPE", "WordPiece", or asking about "train tokenizer", "custom vocabulary", "tokenization", "subword", "fast tokenizer", "encode text"
model-optimization
pluginagentmarketplace
Quantization, pruning, AutoML, hyperparameter tuning, and performance optimization. Use for improving model performance, reducing size, or automated ML.
sakthai-safety-secrets
beer-sakthai
Never capture secrets into memory.
engram
qwexs
Etalon memory architecture with Knowledge Graph (PARA), session isolation, memory decay, and QMD hybrid search
obsidian-knowledge
molotroka123-cell
Use the configured Obsidian vault as a read-mostly local knowledge base, with agent writes restricted to the BOSSMAN Memory folder.
ti-ui
macCesar
"Titanium SDK UI/UX patterns and components expert. Use when working with, reviewing, analyzing, or examining Titanium layouts, ListView/TableView performance optimization, event handling and bubbling, gestures (swipe, pinch), animations, accessibility (VoiceOver/TalkBack), orientation changes, custom fonts/icons, app icons/splash screens, or platform-specific UI (Action Bar, Navigation Bar)."
flutter-performance
spjoshis
Optimize Flutter app performance with widget rebuilds, memory management, rendering optimization, and profiling techniques. Achieve smooth 60fps rendering.
agile-metrics
spjoshis
Master agile metrics with velocity, burn-down charts, cycle time, and team health indicators for data-driven improvement.
memory-curator
molotroka123-cell
Convert completed work into concise durable memory containing decisions, lessons, successful procedures, failures worth remembering and next actions without storing raw transcript noise.
sakthai-memory-search
beer-sakthai
Find specific facts or observations by keyword.
optimizing-attention-flash
magnetoid
Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.
context-degradation
eyadsibai
Use when diagnosing agent failures, debugging lost-in-middle issues, understanding context poisoning, or asking about "context degradation", "lost in middle", "context poisoning", "attention patterns", "context clash", "agent performance drops"
peft-fine-tuning
magnetoid
Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.
sakthai-learning-curation
beer-sakthai
Keep memory healthy over time — consolidate, dedupe, and forget stale facts.
openclaw-workspace-sowork
cj-wang-sowork
Production-ready OpenClaw workspace template for marketing teams. Includes brand positioning, web research, and content writing skills, plus a full AGENTS/SOUL/TOOLS/USER/IDENTITY/HEARTBEAT/MEMORY workspace system. Battle-tested across 13 markets by SoWork. DeepEval score: 0.940.
loudness-metering
SpiralCloudOmega
EBU R128 / ITU-R BS.1770-4 loudness metering, K-weighting, gating, true peak, LRA, and streaming targets for REVITHION STUDIO.
project-memory
abpai
Always-on project memory workflow. At the start of every task, read .agents/LEARNINGS.md; while working, update it with mistakes, corrections, successful and failed patterns, and project-specific preferences.
ai-persona-creator
Tempuss
Use when analyzing stakeholder psychology for negotiations, proposals, or persuasion. Creates research-backed personas revealing hidden motivations.
sakthai-memory-store
beer-sakthai
Write durable memory through the one store seam, choosing the right shape.
query-optimizer
CuriousLearner
Analyze and optimize SQL queries for better performance and efficiency.
performance-profiler
CuriousLearner
Analyze code performance patterns and identify optimization opportunities.
rust-cargo-assistant
CuriousLearner
Cargo build system, crate management, and Rust project configuration assistance.
resource-monitor
CuriousLearner
Monitor system resources (CPU, memory, disk, network) during development and production.