Automates a web browser and captures AI-friendly page snapshots for agents.
ruvnet/RuView
Turns commodity WiFi signals into room-level sensing for presence, breathing, heartbeat and posture, with setup, calibration and training skills.
- 54 skills
- 99.7k stars
- Reached #8 on GitHub's daily trending
The 5 most installed
Ordered by installs on skills.sh- browser49 installs
- Pair Programming44 installs
Pairs with you on coding in driver, navigator or switch modes, with continuous review, testing and security scanning.
- Skill Builder44 installs
Creates new Claude Code skills with correct frontmatter, progressive disclosure and directory layout.
- AgentDB Memory Patterns41 installs
Adds persistent memory to agents with AgentDB, covering session memory, long-term storage and context management.
- Swarm Orchestration41 installs
Orchestrates multi-agent swarms for parallel task execution with dynamic topology and coordination.
The other 49 skills
- V3 Swarm Coordination41 installs
Coordinates 15 agents in a hierarchical mesh to run parallel security, core and integration work.
- github-workflow-automation40 installs
Builds and automates GitHub Actions CI/CD workflows with agent coordination.
- V3 Performance Optimization40 installs
Benchmarks and tunes claude-flow v3 for faster attention, quicker search and lower memory use.
- swarm-advanced39 installs
Applies advanced swarm orchestration patterns to research, development, testing and distributed workflows.
- V3 DDD Architecture39 installs
Builds claude-flow v3 with domain-driven design, bounded contexts and a microkernel pattern.
- github-code-review38 installs
Reviews GitHub pull requests and code with coordinated multi-agent analysis.
- github-multi-repo38 installs
Coordinates and synchronizes work across multiple GitHub repositories with agent orchestration.
- github-release-management38 installs
Orchestrates GitHub releases with versioning, testing, deployment and rollback steps.
- sparc-methodology38 installs
Guides development through the SPARC phases from specification to completion with multi-agent orchestration.
- V3 CLI Modernization38 installs
Modernizes the claude-flow v3 command line with interactive prompts, command splitting and hook integration.
- V3 Memory Unification38 installs
Merges multiple memory systems into AgentDB with fast HNSW-indexed search.
- V3 Security Overhaul38 installs
Reworks claude-flow v3 security to fix known CVEs and apply secure-by-default patterns.
- AgentDB Vector Search37 installs
Implements semantic vector search with AgentDB for document retrieval, similarity matching and RAG systems.
- github-project-management37 installs
Manages GitHub issues, project boards and sprint planning with coordinated agents.
- Hooks Automation37 installs
Sets up Claude Code hooks for task coordination, formatting, session management, git integration and memory.
- stream-chain37 installs
Chains multi-agent pipelines and sequential workflows using streamed JSON data.
- V3 Core Implementation37 installs
Implements claude-flow v3 core modules with clean architecture, dependency injection and tests.
- V3 MCP Optimization37 installs
Optimizes the claude-flow v3 MCP server with connection pooling, load balancing and fast response times.
- ReasoningBank Intelligence36 installs
Adds adaptive learning with ReasoningBank for pattern recognition, strategy optimization and continuous improvement.
- ruview-cli-api26 installs
Uses the RuView command-line tool, REST API and browser build to integrate sensing into other programs.
- ruview-hardware-setup26 installs
Builds, flashes and provisions ESP32 sensing nodes and helps debug devices that stop streaming data.
- ruview-applications25 installs
Runs RuView sensing apps for presence, vitals, fall detection, pose estimation, sleep monitoring and 3D mapping.
- ruview-model-training25 installs
Trains, fine-tunes and publishes RuView pose and embedding models, including GPU training and Hugging Face uploads.
- ruview-advanced-sensing24 installs
Sets up research-grade multi-node sensing with cross-viewpoint fusion, RF tomography and adversarial signal detection.
- ruview-mmwave24 installs
Sets up mmWave radar sensors and fuses their readings with WiFi sensing for vitals and presence.
- ruview-verify24 installs
Runs the full test suite, firmware hash checks and witness bundle to verify a RuView build.
- ruview-configure23 installs
Adjusts a deployed RuView system's settings, including WiFi channels, edge modules, mesh and server flags.
- ruview-quickstart23 installs
Walks a newcomer through the fastest Docker demo and build path to a live sensing dashboard.
- ruview-rvagent20 installs
Prototypes rvAgent and RVF integration so agents can read live sensing events and store state alongside data.
- calibrate-room1 installs
Runs the per-room calibration pipeline that trains small models for presence, posture, breathing, heartbeat and restlessness.
- onboard1 installs
Picks a starting path for RuView, such as docker demo, repo build or live ESP32, and runs the next step.
- provision-node1 installs
Builds, flashes and provisions an ESP32 CSI node for RuView, including firmware choice and WiFi settings.
- train-pose1 installs
Trains and evaluates WiFi pose models, always comparing results against the mean-pose baseline.
- verify1 installs
Runs the SHA-256 proof and witness bundle for a RuView result and flags claims that are not measured.
- verify-homecore1 installs
Runs the smallest relevant Homecore test profile, from core to Wasmtime, HAP or full.
Sets up advanced AgentDB features like QUIC sync, multi-database management, custom distance metrics and hybrid search.
Builds and trains reinforcement learning plugins for AgentDB, including Q-Learning, SARSA and Actor-Critic.
Speeds up AgentDB with quantization, HNSW indexing, caching and batch operations for large vector sets.
Provides guidance for working with Cognitum spaces, based on what the skill name suggests.
Connects to RuView sensing hardware such as ESP32 CSI nodes, mmWave radars and LiDAR, locally or over SSH.
Runs Darwin Mode to evolve the agent harness in a sandboxed, safety-gated way while keeping the model frozen.
Maps a Homecore capability to its reviewed source, tests, architecture decisions and known limitations.
Turns a feature request into a minimal, file-level implementation plan before writing any code.
Combines ReasoningBank learning with AgentDB vector storage for trajectory tracking, verdicts and memory distillation.
Reviews Home Assistant migrations, treating imported data as untrusted and avoiding overwriting existing files.
Reviews Homecore server startup, restore, authentication, feature and provider configuration.
Reviews trust boundaries for native registrations and external Wasm plugins.
Integrates claude-flow v3 deeply with agentic-flow to remove duplicated code.
Scores output truthfulness, checks code quality and rolls back changes that fall below a 0.95 threshold.