agentic-flow 3.1.0 → 3.3.0

package/README.md

@@ -362,6 +362,21 @@ additionally needs the official `swebench` Docker harness — see
 - **Performance**: `benchmark_run`, `bottleneck_analyze`, `token_usage`
 - **And 200+ more tools!**
 
+### 🔌 Ecosystem integrations (3.2.0, solver core in 3.3.0)
+
+Optional ruvnet packages, installed by default and loaded lazily, each with a built-in fallback (`npm install --omit=optional` skips them):
+
+| Package | Powers | MCP tool (readonly profile) |
+|---|---|---|
+| [`aidefence`](https://www.npmjs.com/package/aidefence) | prompt-injection screening of every routed task (`AGENTIC_FLOW_PROMPT_INJECTION`) | `prompt_injection_scan` |
+| [`@ruvector/mincut-wasm`](https://www.npmjs.com/package/@ruvector/mincut-wasm) | split a task-dependency graph between two agent teams (Node >= 22; built-in fallback on 20) | `swarm_partition_tasks` |
+| [`@ruvnet/sublinear-core`](https://www.npmjs.com/package/@ruvnet/sublinear-core) | verified linear solves, PageRank, personalized PageRank and forward-push top-k (zero dependencies) | `sublinear_solve`, `sublinear_pagerank` |
+| [`agent-booster`](https://www.npmjs.com/package/agent-booster) | in-process code edits | `agent_booster_edit_file` (admin) |
+| [`@ruvector/router`](https://www.npmjs.com/package/@ruvector/router) | native HNSW index for `SemanticRouter` | — |
+| [`ruvector-onnx-embeddings-wasm`](https://www.npmjs.com/package/ruvector-onnx-embeddings-wasm) | all-MiniLM-L6-v2 embeddings for `AgentDBFast` | — |
+
+`ecosystem_status` reports what is installed; `import { partitionTaskGraph, pageRank, scanPromptInjection } from 'agentic-flow/integrations'` gives the same functions in code. Skills: `.claude/skills/agentic-flow-ecosystem` (Claude Code) and `.agents/skills/agentic-flow-ecosystem` (Codex), both installed by `agentic-flow init`.
+
 ### 🧩 Advanced Capabilities
 
 - **🧠 ReasoningBank Learning Memory**: All 66 agents learn from every task execution
@@ -1050,16 +1065,16 @@ The `init` command sets up your project with the full Agentic-Flow infrastructur
 
 ```bash
 # Initialize project with full agent library
-npx agentic-flow@3.1.0 init
+npx agentic-flow@3.2.0 init
 
 # Force reinitialize (overwrite existing)
-npx agentic-flow@3.1.0 init --force
+npx agentic-flow@3.2.0 init --force
 
 # Minimal setup (empty directories only)
-npx agentic-flow@3.1.0 init --minimal
+npx agentic-flow@3.2.0 init --minimal
 
 # Verbose output showing all files
-npx agentic-flow@3.1.0 init --verbose
+npx agentic-flow@3.2.0 init --verbose
 ```
 
 ### What Gets Created
@@ -1103,7 +1118,7 @@ The generated `settings.json` includes:
   "mcpServers": {
     "claude-flow": {
       "command": "npx",
-      "args": ["agentic-flow@3.1.0", "mcp", "start"]
+      "args": ["agentic-flow@3.2.0", "mcp", "start"]
     }
   }
 }
@@ -1115,13 +1130,13 @@ After initialization:
 
 ```bash
 # 1. Start the MCP server
-npx agentic-flow@3.1.0 mcp start
+npx agentic-flow@3.2.0 mcp start
 
 # 2. Bootstrap intelligence from your codebase
-npx agentic-flow@3.1.0 hooks pretrain
+npx agentic-flow@3.2.0 hooks pretrain
 
 # 3. Generate optimized agent configurations
-npx agentic-flow@3.1.0 hooks build-agents
+npx agentic-flow@3.2.0 hooks build-agents
 
 # 4. Start using Claude Code
 claude
@@ -1155,14 +1170,14 @@ Agentic-Flow v2 includes a powerful **self-learning hooks system** powered by Ru
 Get context and agent suggestions before editing a file:
 
 ```bash
-npx agentic-flow@3.1.0 hooks pre-edit <filePath> [options]
+npx agentic-flow@3.2.0 hooks pre-edit <filePath> [options]
 
 Options:
   -t, --task <task>   Task description
   -j, --json          Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks pre-edit src/api/users.ts --task "Add validation"
+npx agentic-flow@3.2.0 hooks pre-edit src/api/users.ts --task "Add validation"
 # Output:
 # 🎯 Suggested Agent: backend-dev
 # 📊 Confidence: 94.2%
@@ -1177,7 +1192,7 @@ npx agentic-flow@3.1.0 hooks pre-edit src/api/users.ts --task "Add validation"
 Record edit outcome for learning:
 
 ```bash
-npx agentic-flow@3.1.0 hooks post-edit <filePath> [options]
+npx agentic-flow@3.2.0 hooks post-edit <filePath> [options]
 
 Options:
   -s, --success           Mark as successful edit
@@ -1188,10 +1203,10 @@ Options:
   -j, --json              Output as JSON
 
 # Example (success)
-npx agentic-flow@3.1.0 hooks post-edit src/api/users.ts --success --agent coder
+npx agentic-flow@3.2.0 hooks post-edit src/api/users.ts --success --agent coder
 
 # Example (failure)
-npx agentic-flow@3.1.0 hooks post-edit src/api/users.ts --fail --error "Type error"
+npx agentic-flow@3.2.0 hooks post-edit src/api/users.ts --fail --error "Type error"
 ```
 
 #### Pre-Command Hook
@@ -1199,13 +1214,13 @@ npx agentic-flow@3.1.0 hooks post-edit src/api/users.ts --fail --error "Type err
 Assess command risk before execution:
 
 ```bash
-npx agentic-flow@3.1.0 hooks pre-command "<command>" [options]
+npx agentic-flow@3.2.0 hooks pre-command "<command>" [options]
 
 Options:
   -j, --json    Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks pre-command "rm -rf node_modules"
+npx agentic-flow@3.2.0 hooks pre-command "rm -rf node_modules"
 # Output:
 # ⚠️ Risk Level: CAUTION (65%)
 # ✅ Command APPROVED
@@ -1218,7 +1233,7 @@ npx agentic-flow@3.1.0 hooks pre-command "rm -rf node_modules"
 Route task to optimal agent using learned patterns:
 
 ```bash
-npx agentic-flow@3.1.0 hooks route "<task>" [options]
+npx agentic-flow@3.2.0 hooks route "<task>" [options]
 
 Options:
   -f, --file <filePath>   Context file path
@@ -1226,7 +1241,7 @@ Options:
   -j, --json              Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks route "Fix authentication bug in login flow"
+npx agentic-flow@3.2.0 hooks route "Fix authentication bug in login flow"
 # Output:
 # 🎯 Recommended Agent: backend-dev
 # 📊 Confidence: 91.5%
@@ -1245,14 +1260,14 @@ npx agentic-flow@3.1.0 hooks route "Fix authentication bug in login flow"
 Explain routing decision with full transparency:
 
 ```bash
-npx agentic-flow@3.1.0 hooks explain "<task>" [options]
+npx agentic-flow@3.2.0 hooks explain "<task>" [options]
 
 Options:
   -f, --file <filePath>   Context file path
   -j, --json              Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks explain "Implement caching layer"
+npx agentic-flow@3.2.0 hooks explain "Implement caching layer"
 # Output:
 # 📝 Summary: Task involves performance optimization and data caching
 # 🎯 Recommended: perf-analyzer
@@ -1273,7 +1288,7 @@ npx agentic-flow@3.1.0 hooks explain "Implement caching layer"
 Analyze repository to bootstrap intelligence:
 
 ```bash
-npx agentic-flow@3.1.0 hooks pretrain [options]
+npx agentic-flow@3.2.0 hooks pretrain [options]
 
 Options:
   -d, --depth <n>     Git history depth (default: 50)
@@ -1282,7 +1297,7 @@ Options:
   -j, --json          Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks pretrain --depth 100
+npx agentic-flow@3.2.0 hooks pretrain --depth 100
 # Output:
 # 🧠 Analyzing repository...
 # 📊 Pretrain Complete!
@@ -1299,7 +1314,7 @@ npx agentic-flow@3.1.0 hooks pretrain --depth 100
 Generate optimized agent configurations from pretrain data:
 
 ```bash
-npx agentic-flow@3.1.0 hooks build-agents [options]
+npx agentic-flow@3.2.0 hooks build-agents [options]
 
 Options:
   -f, --focus <mode>    Focus: quality|speed|security|testing|fullstack
@@ -1309,7 +1324,7 @@ Options:
   -j, --json            Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks build-agents --focus security
+npx agentic-flow@3.2.0 hooks build-agents --focus security
 # Output:
 # ✅ Agents Generated!
 #    📦 Total: 12
@@ -1327,7 +1342,7 @@ npx agentic-flow@3.1.0 hooks build-agents --focus security
 View learning metrics and performance dashboard:
 
 ```bash
-npx agentic-flow@3.1.0 hooks metrics [options]
+npx agentic-flow@3.2.0 hooks metrics [options]
 
 Options:
   -t, --timeframe <period>   Timeframe: 1h|24h|7d|30d (default: 24h)
@@ -1335,7 +1350,7 @@ Options:
   -j, --json                 Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks metrics --timeframe 7d --detailed
+npx agentic-flow@3.2.0 hooks metrics --timeframe 7d --detailed
 # Output:
 # 📊 Learning Metrics (7d)
 #
@@ -1357,7 +1372,7 @@ npx agentic-flow@3.1.0 hooks metrics --timeframe 7d --detailed
 Transfer learned patterns from another project:
 
 ```bash
-npx agentic-flow@3.1.0 hooks transfer <sourceProject> [options]
+npx agentic-flow@3.2.0 hooks transfer <sourceProject> [options]
 
 Options:
   -c, --min-confidence <n>   Minimum confidence threshold (default: 0.7)
@@ -1366,7 +1381,7 @@ Options:
   -j, --json                 Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks transfer ../other-project --mode merge
+npx agentic-flow@3.2.0 hooks transfer ../other-project --mode merge
 # Output:
 # ✅ Transfer Complete!
 #    📥 Patterns transferred: 45
@@ -1384,7 +1399,7 @@ The `intelligence` (alias: `intel`) subcommand provides access to the full RuVec
 Route task using SONA + MoE + HNSW (150x faster than brute force):
 
 ```bash
-npx agentic-flow@3.1.0 hooks intelligence route "<task>" [options]
+npx agentic-flow@3.2.0 hooks intelligence route "<task>" [options]
 
 Options:
   -f, --file <path>       File context
@@ -1393,7 +1408,7 @@ Options:
   -j, --json              Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 hooks intel route "Optimize database queries" --top-k 3
+npx agentic-flow@3.2.0 hooks intel route "Optimize database queries" --top-k 3
 # Output:
 # ⚡ RuVector Intelligence Route
 # 🎯 Agent: perf-analyzer
@@ -1409,15 +1424,15 @@ Track reinforcement learning trajectories for agent improvement:
 
 ```bash
 # Start a trajectory
-npx agentic-flow@3.1.0 hooks intel trajectory-start "<task>" -a <agent>
+npx agentic-flow@3.2.0 hooks intel trajectory-start "<task>" -a <agent>
 # Output: 🎬 Trajectory Started - ID: 42
 
 # Record steps
-npx agentic-flow@3.1.0 hooks intel trajectory-step 42 -a "edit file" -r 0.8
-npx agentic-flow@3.1.0 hooks intel trajectory-step 42 -a "run tests" -r 1.0 --test-passed
+npx agentic-flow@3.2.0 hooks intel trajectory-step 42 -a "edit file" -r 0.8
+npx agentic-flow@3.2.0 hooks intel trajectory-step 42 -a "run tests" -r 1.0 --test-passed
 
 # End trajectory
-npx agentic-flow@3.1.0 hooks intel trajectory-end 42 --success --quality 0.95
+npx agentic-flow@3.2.0 hooks intel trajectory-end 42 --success --quality 0.95
 # Output: 🏁 Trajectory Completed - Learning: EWC++ consolidation applied
 ```
 
@@ -1427,13 +1442,13 @@ Store and search patterns using HNSW-indexed ReasoningBank:
 
 ```bash
 # Store a pattern
-npx agentic-flow@3.1.0 hooks intel pattern-store \
+npx agentic-flow@3.2.0 hooks intel pattern-store \
   --task "Fix React hydration error" \
   --resolution "Use useEffect with empty deps for client-only code" \
   --score 0.95
 
 # Search patterns (150x faster with HNSW)
-npx agentic-flow@3.1.0 hooks intel pattern-search "hydration mismatch"
+npx agentic-flow@3.2.0 hooks intel pattern-search "hydration mismatch"
 # Output:
 # 🔍 Pattern Search Results
 #    Query: "hydration mismatch"
@@ -1449,7 +1464,7 @@ npx agentic-flow@3.1.0 hooks intel pattern-search "hydration mismatch"
 Get RuVector intelligence layer statistics:
 
 ```bash
-npx agentic-flow@3.1.0 hooks intelligence stats
+npx agentic-flow@3.2.0 hooks intelligence stats
 # Output:
 # 📊 RuVector Intelligence Stats
 #
@@ -1490,7 +1505,7 @@ The `init` command automatically configures hooks in `.claude/settings.json`:
         "hooks": [
           {
             "type": "command",
-            "command": "npx agentic-flow@3.1.0 hooks pre-edit \"$TOOL_INPUT_file_path\""
+            "command": "npx agentic-flow@3.2.0 hooks pre-edit \"$TOOL_INPUT_file_path\""
           }
         ]
       },
@@ -1499,7 +1514,7 @@ The `init` command automatically configures hooks in `.claude/settings.json`:
         "hooks": [
           {
             "type": "command",
-            "command": "npx agentic-flow@3.1.0 hooks pre-command \"$TOOL_INPUT_command\""
+            "command": "npx agentic-flow@3.2.0 hooks pre-command \"$TOOL_INPUT_command\""
           }
         ]
       }
@@ -1510,7 +1525,7 @@ The `init` command automatically configures hooks in `.claude/settings.json`:
         "hooks": [
           {
             "type": "command",
-            "command": "npx agentic-flow@3.1.0 hooks post-edit \"$TOOL_INPUT_file_path\" --success"
+            "command": "npx agentic-flow@3.2.0 hooks post-edit \"$TOOL_INPUT_file_path\" --success"
           }
         ]
       }
@@ -1521,7 +1536,7 @@ The `init` command automatically configures hooks in `.claude/settings.json`:
         "hooks": [
           {
             "type": "command",
-            "command": "npx agentic-flow@3.1.0 hooks post-edit \"$TOOL_INPUT_file_path\" --fail --error \"$ERROR_MESSAGE\""
+            "command": "npx agentic-flow@3.2.0 hooks post-edit \"$TOOL_INPUT_file_path\" --fail --error \"$ERROR_MESSAGE\""
           }
         ]
       }
@@ -1531,7 +1546,7 @@ The `init` command automatically configures hooks in `.claude/settings.json`:
         "hooks": [
           {
             "type": "command",
-            "command": "npx agentic-flow@3.1.0 hooks intelligence stats --json"
+            "command": "npx agentic-flow@3.2.0 hooks intelligence stats --json"
           }
         ]
       }
@@ -1542,7 +1557,7 @@ The `init` command automatically configures hooks in `.claude/settings.json`:
           {
             "type": "command",
             "timeout": 3000,
-            "command": "npx agentic-flow@3.1.0 hooks route \"$USER_PROMPT\" --json"
+            "command": "npx agentic-flow@3.2.0 hooks route \"$USER_PROMPT\" --json"
           }
         ]
       }
@@ -1608,10 +1623,10 @@ Workers are automatically dispatched when trigger keywords are detected in promp
 Detect triggers in prompt and dispatch background workers:
 
 ```bash
-npx agentic-flow@3.1.0 workers dispatch "<prompt>"
+npx agentic-flow@3.2.0 workers dispatch "<prompt>"
 
 # Example
-npx agentic-flow@3.1.0 workers dispatch "ultralearn how authentication works"
+npx agentic-flow@3.2.0 workers dispatch "ultralearn how authentication works"
 # Output:
 # ⚡ Background Workers Spawned:
 #   • ultralearn: worker-1234
@@ -1624,7 +1639,7 @@ npx agentic-flow@3.1.0 workers dispatch "ultralearn how authentication works"
 Get worker status and progress:
 
 ```bash
-npx agentic-flow@3.1.0 workers status [workerId]
+npx agentic-flow@3.2.0 workers status [workerId]
 
 Options:
   -s, --session <id>   Filter by session
@@ -1632,7 +1647,7 @@ Options:
   -j, --json           Output as JSON
 
 # Example - Dashboard view
-npx agentic-flow@3.1.0 workers status
+npx agentic-flow@3.2.0 workers status
 # Output:
 # ┌─ Background Workers Dashboard ────────────┐
 # │ ✅ ultralearn: complete                    │
@@ -1650,7 +1665,7 @@ npx agentic-flow@3.1.0 workers status
 View worker analysis results:
 
 ```bash
-npx agentic-flow@3.1.0 workers results [workerId]
+npx agentic-flow@3.2.0 workers results [workerId]
 
 Options:
   -s, --session <id>    Filter by session
@@ -1658,7 +1673,7 @@ Options:
   -j, --json            Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 workers results
+npx agentic-flow@3.2.0 workers results
 # Output:
 # 📊 Worker Analysis Results
 #   • ultralearn "authentication":
@@ -1674,7 +1689,7 @@ npx agentic-flow@3.1.0 workers results
 List all available trigger keywords:
 
 ```bash
-npx agentic-flow@3.1.0 workers triggers
+npx agentic-flow@3.2.0 workers triggers
 # Output:
 # ⚡ Available Background Worker Triggers:
 # ┌──────────────┬──────────┬────────────────────────────────────────┐
@@ -1692,14 +1707,14 @@ npx agentic-flow@3.1.0 workers triggers
 Get worker statistics:
 
 ```bash
-npx agentic-flow@3.1.0 workers stats [options]
+npx agentic-flow@3.2.0 workers stats [options]
 
 Options:
   -t, --timeframe <period>   Timeframe: 1h, 24h, 7d (default: 24h)
   -j, --json                 Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 workers stats --timeframe 7d
+npx agentic-flow@3.2.0 workers stats --timeframe 7d
 # Output:
 # ⚡ Worker Statistics (7d)
 # Total Workers: 45
@@ -1723,14 +1738,14 @@ Create and manage custom workers with specific analysis phases:
 #### List Presets
 
 ```bash
-npx agentic-flow@3.1.0 workers presets
+npx agentic-flow@3.2.0 workers presets
 # Shows available worker presets: quick-scan, deep-analysis, security-audit, etc.
 ```
 
 #### Create Custom Worker
 
 ```bash
-npx agentic-flow@3.1.0 workers create <name> [options]
+npx agentic-flow@3.2.0 workers create <name> [options]
 
 Options:
   -p, --preset <preset>     Preset to use (default: quick-scan)
@@ -1738,13 +1753,13 @@ Options:
   -d, --description <desc>  Worker description
 
 # Example
-npx agentic-flow@3.1.0 workers create security-check --preset security-audit --triggers "security,vuln"
+npx agentic-flow@3.2.0 workers create security-check --preset security-audit --triggers "security,vuln"
 ```
 
 #### Run Custom Worker
 
 ```bash
-npx agentic-flow@3.1.0 workers run <nameOrTrigger> [options]
+npx agentic-flow@3.2.0 workers run <nameOrTrigger> [options]
 
 Options:
   -t, --topic <topic>    Topic to analyze
@@ -1752,7 +1767,7 @@ Options:
   -j, --json             Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 workers run security-check --topic "authentication flow"
+npx agentic-flow@3.2.0 workers run security-check --topic "authentication flow"
 ```
 
 ### Native RuVector Workers
@@ -1760,7 +1775,7 @@ npx agentic-flow@3.1.0 workers run security-check --topic "authentication flow"
 Run native RuVector workers for advanced analysis:
 
 ```bash
-npx agentic-flow@3.1.0 workers native <type> [options]
+npx agentic-flow@3.2.0 workers native <type> [options]
 
 Types:
   security   - Run security vulnerability scan
@@ -1769,7 +1784,7 @@ Types:
   phases     - List available native phases
 
 # Example
-npx agentic-flow@3.1.0 workers native security
+npx agentic-flow@3.2.0 workers native security
 # Output:
 # ⚡ Native Worker: security
 # ══════════════════════════════════════════════════
@@ -1796,7 +1811,7 @@ npx agentic-flow@3.1.0 workers native security
 Run performance benchmarks on the worker system:
 
 ```bash
-npx agentic-flow@3.1.0 workers benchmark [options]
+npx agentic-flow@3.2.0 workers benchmark [options]
 
 Options:
   -t, --type <type>         Benchmark type: all, trigger-detection, registry,
@@ -1805,7 +1820,7 @@ Options:
   -j, --json                Output as JSON
 
 # Example
-npx agentic-flow@3.1.0 workers benchmark --type trigger-detection
+npx agentic-flow@3.2.0 workers benchmark --type trigger-detection
 # Output:
 # ✅ Trigger Detection Benchmark
 #    Operation: detect triggers in prompts
@@ -1820,7 +1835,7 @@ npx agentic-flow@3.1.0 workers benchmark --type trigger-detection
 View worker-agent integration statistics:
 
 ```bash
-npx agentic-flow@3.1.0 workers integration
+npx agentic-flow@3.2.0 workers integration
 # Output:
 # ⚡ Worker-Agent Integration Stats
 # ════════════════════════════════════════
@@ -1841,10 +1856,10 @@ npx agentic-flow@3.1.0 workers integration
 Get recommended agents for a worker trigger:
 
 ```bash
-npx agentic-flow@3.1.0 workers agents <trigger>
+npx agentic-flow@3.2.0 workers agents <trigger>
 
 # Example
-npx agentic-flow@3.1.0 workers agents ultralearn
+npx agentic-flow@3.2.0 workers agents ultralearn
 # Output:
 # ⚡ Agent Recommendations for "ultralearn"
 #
@@ -1873,7 +1888,7 @@ Workers are automatically configured in `.claude/settings.json` via hooks:
             "type": "command",
             "timeout": 5000,
             "background": true,
-            "command": "npx agentic-flow@3.1.0 workers dispatch-prompt \"$USER_PROMPT\" --session \"$SESSION_ID\" --json"
+            "command": "npx agentic-flow@3.2.0 workers dispatch-prompt \"$USER_PROMPT\" --session \"$SESSION_ID\" --json"
           }
         ]
       }
@@ -1883,7 +1898,7 @@ Workers are automatically configured in `.claude/settings.json` via hooks:
         "hooks": [
           {
             "type": "command",
-            "command": "npx agentic-flow@3.1.0 workers cleanup --age 24"
+            "command": "npx agentic-flow@3.2.0 workers cleanup --age 24"
           }
         ]
       }
@@ -1906,7 +1921,7 @@ Workers are automatically configured in `.claude/settings.json` via hooks:
 
 ```bash
 # Install the current release
-npm install agentic-flow@3.1.0
+npm install agentic-flow@3.2.0
 ```
 
 ### Install from Source
@@ -2226,7 +2241,7 @@ MIT License - see [LICENSE](LICENSE) file for details.
 **Install now and experience the future of AI agents:**
 
 ```bash
-npm install agentic-flow@3.1.0
+npm install agentic-flow@3.2.0
 ```
 
 **Made with ❤️ by [@ruvnet](https://github.com/ruvnet)**