admin/ruflo

0 Fork 0

mirror of https://github.com/ruvnet/ruflo.git synced 2026-10-28 15:48:10 +09:00

Table of Contents

MCP Tools

Tool Categories

Swarm Coordination (40+ tools)
Agent Management (50+ tools)
Memory & Graph Intelligence (80+ tools)
Hooks & Learning (40+ tools)
Security & Safety (25+ tools)
Task & Workflow (30+ tools)
GitHub Integration (35+ tools)
Intelligence & Neural (35+ tools)
Browser Automation (20+ tools)
Performance & Observability (30+ tools)
Cost Tracking (15+ tools)
Advanced Features (ADR-129, ADR-130)

Quick Reference by Use Case
Usage Example

MCP Tools

Ruflo exposes 314 MCP tools for Claude Code to call directly. These are organized by domain.

Tool Categories

Swarm Coordination (40+ tools)

swarm_init, swarm_status, swarm_shutdown, swarm_health
coordination_consensus, coordination_load_balance, coordination_metrics
hive-mind_init, hive-mind_join, hive-mind_memory, hive-mind_consensus

Agent Management (50+ tools)

agent_spawn, agent_list, agent_status, agent_execute, agent_terminate
agent_health, agent_update, agent_logs
claims_claim, claims_accept-handoff, claims_mark-stealable, claims_steal

Memory & Graph Intelligence (80+ tools)

Flat Memory:

memory_store, memory_retrieve, memory_search, memory_list, memory_delete
memory_export, memory_import, memory_migrate, memory_compress, memory_stats

AgentDB Controllers (ADR-130):

agentdb_batch — insert, update, delete episodes
agentdb_causal-edge — record causal links (A causes B, A depends on B, A supersedes B)
agentdb_causal-edge-delete, agentdb_causal-node-delete — clean up stale edges
agentdb_hierarchical-store, agentdb_hierarchical-recall, agentdb_hierarchical-delete
agentdb_pattern-search, agentdb_pattern-store
agentdb_semantic-route, agentdb_route
agentdb_context-synthesize
agentdb_session-start, agentdb_session-end
agentdb_feedback, agentdb_learning_status
agentdb_health, agentdb_controllers

Vector Search (HNSW + RaBitQ):

embeddings_generate — text to vector
embeddings_search, embeddings_search_text, embeddings_search_text_batch
embeddings_search_text_diverse — MMR ranking
embeddings_search_text_ensemble — RRF fusion
embeddings_search_text_hyde — hypothetical answer embedding
embeddings_ann_router_build, embeddings_ann_router_search, embeddings_ann_router_status
embeddings_diskann_build, embeddings_diskann_search, embeddings_diskann_status
embeddings_rabitq_build, embeddings_rabitq_search, embeddings_rabitq_status
embeddings_compare, embeddings_hyperbolic, embeddings_neural
embeddings_init, embeddings_status

Hooks & Learning (40+ tools)

hooks_route, hooks_explain, hooks_intelligence, hooks_intelligence_pattern-store, hooks_intelligence_pattern-search
hooks_pre-task, hooks_post-task, hooks_pre-edit, hooks_post-edit
hooks_session-start, hooks_session-end, hooks_session-restore
hooks_worker-list, hooks_worker-dispatch, hooks_worker-status, hooks_worker-detect, hooks_worker-cancel
hooks_intelligence_trajectory-start, hooks_intelligence_trajectory-step, hooks_intelligence_trajectory-end

Security & Safety (25+ tools)

aidefence_scan, aidefence_has_pii, aidefence_is_safe, aidefence_analyze
aidefence_learn, aidefence_stats

Task & Workflow (30+ tools)

task_create, task_list, task_status, task_update, task_assign, task_complete, task_cancel, task_retry, task_summary
workflow_create, workflow_list, workflow_status, workflow_execute, workflow_pause, workflow_resume, workflow_stop, workflow_cancel, workflow_delete
workflow_template, workflow_validate, workflow_run

GitHub Integration (35+ tools)

github_pr_manage — create, review, merge, close PRs
github_issue_track — create, update, close issues
github_workflow — trigger, status, cancel GitHub Actions
github_repo_analyze, github_repo_architect
github_metrics

Intelligence & Neural (35+ tools)

neural_train, neural_status, neural_patterns, neural_predict, neural_optimize, neural_compress
hooks_model-route, hooks_model-outcome, hooks_model-stats
daa_agent_create, daa_agent_adapt, daa_cognitive_pattern
daa_knowledge_share, daa_learning_status, daa_performance_metrics
daa_workflow_create, daa_workflow_execute

Browser Automation (20+ tools)

browser_session_record, browser_session_replay, browser_session_end
browser_cookie_use, browser_template_apply

Performance & Observability (30+ tools)

performance_benchmark, performance_profile, performance_metrics, performance_bottleneck, performance_optimize, performance_report
system_status, system_health, system_metrics, system_info, system_reset

Cost Tracking (15+ tools)

cost-track, cost-federation, cost-report, cost-summary, cost-trend
cost-booster-route, cost-booster-edit
cost-optimize, cost-benchmark, cost-budget-check

Advanced Features (ADR-129, ADR-130)

WASM Agents (rvagent): wasm_agent_create, wasm_agent_prompt, wasm_agent_tool, wasm_agent_export, wasm_agent_files, wasm_agent_list, wasm_agent_terminate
Managed Agents (cloud): managed_agent_create, managed_agent_prompt, managed_agent_status, managed_agent_events, managed_agent_list, managed_agent_terminate
Graph Pathfinding (ADR-130): Shortest path, PageRank, influence scores, causal chains
RuVLLM (local inference): ruvllm_chat_format, ruvllm_generate_config, ruvllm_hnsw_create, ruvllm_hnsw_route, ruvllm_hnsw_add
RuVector GPU: vector_embed, vector_cluster, vector_hyperbolic, vector_setup

Quick Reference by Use Case

Use Case	Tools
Create a team	`swarm_init`, `agent_spawn` (multiple)
Persistent memory	`memory_store`, `memory_search`, `memory_retrieve`
AI-enhanced routing	`hooks_route`, `neural_predict`
Verify artifacts	`witness_verify`, `witness_sign`
Find security bugs	`security_scan`, `aidefence_scan`
Optimize performance	`performance_benchmark`, `performance_optimize`
Run tests	`task_create`, `testgen` (coverage analysis)
Track costs	`cost-track`, `cost-report`
Deep learning	`daa_agent_create`, `daa_workflow_execute`, `neural_train`
Local inference	`ruvllm_`, `wasm_agent_`

Usage Example

// Store a memory pattern
await mcp.memory_store({
  key: "fast-sorting",
  value: "QuickSort beats Merge on small arrays (<1000 elems)",
  namespace: "patterns"
})

// Search for related patterns
const results = await mcp.memory_search({
  query: "sorting algorithm performance",
  namespace: "patterns",
  limit: 5
})

// Route a complex task
const recommendation = await mcp.hooks_route({
  task: "Refactor legacy authentication module"
})
// Returns: { model: "sonnet", reason: "Complex refactor across 3 modules" }

For details on any tool, call:

npx ruflo@latest mcp list --tool-name "<tool-name>"