🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
TypeScript agentic-aiagentic-frameworkagentic-workflowagentsai-agentsai-assistantai-skillsautonomous-agentsclaudecodexdsh-pluginharnessmcp-servermulti-agentmulti-agent-systemsnpmskillsswarmswarm-intelligencetypescript
Checkpoint: File edits
rUv3b949120f5ad
diff --git a/AGENTS.md b/AGENTS.mdindex 1425f767b5..62859091b0 100644--- a/AGENTS.md+++ b/AGENTS.md@@ -511,55 +511,85 @@ Once added, Codex can use these tools via MCP:| `task_orchestrate` | Coordinate tasks || `neural_train` | Train on successful patterns |-### Learning via Memory (IMPORTANT - USE THIS!)+### Self-Learning via MCP Tools (PREFERRED)-**BEFORE starting any task:**-```bash-# Search for relevant patterns from past work-npx claude-flow memory search --query "keywords related to your task"+Use MCP tools directly - faster than CLI commands:-# Example: before building an API-npx claude-flow memory search --query "API implementation patterns REST endpoints"+**BEFORE starting any task - SEARCH for patterns:**+```+Use tool: memory_search+ query: "keywords related to your task"+ namespace: "patterns"```-If patterns are found, USE THEM to guide your implementation.-**AFTER completing a task successfully:**-```bash-# Store the successful pattern for future reference-npx claude-flow memory store \- --key "pattern-[descriptive-name]" \- --value "Detailed description: what the task was, what approach worked, key code patterns, gotchas to avoid" \- --namespace patterns+**AFTER completing successfully - STORE the pattern:**+```+Use tool: memory_store+ key: "pattern-[descriptive-name]"+ value: "What worked: approach, code patterns, gotchas"+ namespace: "patterns"+```++### MCP Learning Workflow (Use This!)-# Example: after building a worker pool-npx claude-flow memory store \- --key "pattern-concurrent-workers" \- --value "Use bash background jobs: for i in 1..N; do (work) & done; wait. Each worker in subshell for isolation." \- --namespace patterns```+1. LEARN: memory_search(query="task keywords", namespace="patterns")+ → If score > 0.7, USE that pattern-### Self-Learning Workflow (Use This Pattern!)+2. COORDINATE: swarm_init(topology="hierarchical")+ → agent_spawn(type="coder", name="worker-1")-```bash-# 1. LEARN: Search before you start-npx claude-flow memory search --query "task keywords" --namespace patterns+3. EXECUTE: YOU write the code, run commands, create files++4. REMEMBER: memory_store(key="pattern-x", value="what worked", namespace="patterns")+```-# 2. EXECUTE: Do the work (YOU implement)-# ... your implementation here ...+### MCP Tools for Learning-# 3. REMEMBER: Store what worked-npx claude-flow memory store --key "pattern-xxx" --value "what worked" --namespace patterns+| Tool | Purpose | When to Use |+|------|---------|-------------|+| `memory_search` | Find similar past patterns | BEFORE starting any task |+| `memory_store` | Save successful patterns | AFTER completing a task |+| `memory_retrieve` | Get specific pattern by key | When you know the exact key |+| `neural_train` | Train on successful patterns | After multiple successes |++### Example: Learning-Enabled Task-# 4. VERIFY: Confirm pattern is searchable-npx claude-flow memory search --query "related terms" --namespace patterns+```+STEP 1 - LEARN:+Use tool: memory_search+ query: "validation utility function"+ namespace: "patterns"++→ Found: pattern-email-validator (score: 0.82)+→ Use this pattern as reference!++STEP 2 - COORDINATE:+Use tool: swarm_init with topology="hierarchical", maxAgents=3++STEP 3 - EXECUTE:+YOU create the files:+ echo 'export function validate(x) { ... }' > /tmp/validator.js+ node --test /tmp/validator.js++STEP 4 - REMEMBER:+Use tool: memory_store+ key: "pattern-phone-validator"+ value: "Phone validation: regex /^\+?[\d\s-]{10,}$/, normalize first, test edge cases"+ namespace: "patterns"```### Vector Search Tips-- Searches are SEMANTIC (meaning-based, not keyword-based)-- Use descriptive queries: "concurrent worker execution" not just "workers"+- Searches are SEMANTIC (meaning-based, not just keywords)- Score > 0.7 = strong match, use that pattern- Score 0.5-0.7 = partial match, adapt as needed-- Store with detailed values for better retrieval+- Store DETAILED values for better future retrieval++### CLI Fallback (if MCP unavailable)+```bash+npx claude-flow memory search --query "keywords" --namespace patterns+npx claude-flow memory store --key "pattern-x" --value "what worked" --namespace patterns+```### Coordination via MCP