reasoningbank-intelligence

Name: reasoningbank-intelligence
Author: spencermarx

301stars24forks

How to install reasoningbank-intelligence?

spencermarx/open-code-review/reasoningbank-intelligence

$npx -y skills add spencermarx/open-code-review --skill reasoningbank-intelligence

Installs into the current project.

Summary

Implement adaptive learning with ReasoningBank for pattern recognition, strategy optimization, and continuous improvement. Use when building self-learning agents, optimizing workflows, or implementing meta-cognitive systems.

Files · 1

View on GitHub

SKILL.md

1	# ReasoningBank Intelligence
2
3	## What This Skill Does
4
5	Implements ReasoningBank's adaptive learning system for AI agents to learn from experience, recognize patterns, and optimize strategies over time. Enables meta-cognitive capabilities and continuous improvement.
6
7	## Prerequisites
8
9	- agentic-flow v3.0.0-alpha.1+
10	- AgentDB v3.0.0-alpha.10+ (for persistence)
11	- Node.js 18+
12
13	## Quick Start
14
15	```typescript
16	import { ReasoningBank } from 'agentic-flow/reasoningbank';
17
18	// Initialize ReasoningBank
19	const rb = new ReasoningBank({
20	persist: true,
21	learningRate: 0.1,
22	adapter: 'agentdb' // Use AgentDB for storage
23	});
24
25	// Record task outcome
26	await rb.recordExperience({
27	task: 'code_review',
28	approach: 'static_analysis_first',
29	outcome: {
30	success: true,
31	metrics: {
32	bugs_found: 5,
33	time_taken: 120,
34	false_positives: 1
35	}
36	},
37	context: {
38	language: 'typescript',
39	complexity: 'medium'
40	}
41	});
42
43	// Get optimal strategy
44	const strategy = await rb.recommendStrategy('code_review', {
45	language: 'typescript',
46	complexity: 'high'
47	});
48	```
49
50	## Core Features
51
52	### 1. Pattern Recognition
53	```typescript
54	// Learn patterns from data
55	await rb.learnPattern({
56	pattern: 'api_errors_increase_after_deploy',
57	triggers: ['deployment', 'traffic_spike'],
58	actions: ['rollback', 'scale_up'],
59	confidence: 0.85
60	});
61
62	// Match patterns
63	const matches = await rb.matchPatterns(currentSituation);
64	```
65
66	### 2. Strategy Optimization
67	```typescript
68	// Compare strategies
69	const comparison = await rb.compareStrategies('bug_fixing', [
70	'tdd_approach',
71	'debug_first',
72	'reproduce_then_fix'
73	]);
74
75	// Get best strategy
76	const best = comparison.strategies[0];
77	console.log(`Best: ${best.name} (score: ${best.score})`);
78	```
79
80	### 3. Continuous Learning
81	```typescript
82	// Enable auto-learning from all tasks
83	await rb.enableAutoLearning({
84	threshold: 0.7, // Only learn from high-confidence outcomes
85	updateFrequency: 100 // Update models every 100 experiences
86	});
87	```
88
89	## Advanced Usage
90
91	### Meta-Learning
92	```typescript
93	// Learn about learning
94	await rb.metaLearn({
95	observation: 'parallel_execution_faster_for_independent_tasks',
96	confidence: 0.95,
97	applicability: {
98	task_types: ['batch_processing', 'data_transformation'],
99	conditions: ['tasks_independent', 'io_bound']
100	}
101	});
102	```
103
104	### Transfer Learning
105	```typescript
106	// Apply knowledge from one domain to another
107	await rb.transferKnowledge({
108	from: 'code_review_javascript',
109	to: 'code_review_typescript',
110	similarity: 0.8
111	});
112	```
113
114	### Adaptive Agents
115	```typescript
116	// Create self-improving agent
117	class AdaptiveAgent {
118	async execute(task: Task) {
119	// Get optimal strategy
120	const strategy = await rb.recommendStrategy(task.type, task.context);
121
122	// Execute with strategy
123	const result = await this.executeWithStrategy(task, strategy);
124
125	// Learn from outcome
126	await rb.recordExperience({
127	task: task.type,
128	approach: strategy.name,
129	outcome: result,
130	context: task.context
131	});
132
133	return result;
134	}
135	}
136	```
137
138	## Integration with AgentDB
139
140	```typescript
141	// Persist ReasoningBank data
142	await rb.configure({
143	storage: {
144	type: 'agentdb',
145	options: {
146	database: './reasoning-bank.db',
147	enableVectorSearch: true
148	}
149	}
150	});
151
152	// Query learned patterns
153	const patterns = await rb.query({
154	category: 'optimization',
155	minConfidence: 0.8,
156	timeRange: { last: '30d' }
157	});
158	```
159
160	## Performance Metrics
161
162	```typescript
163	// Track learning effectiveness
164	const metrics = await rb.getMetrics();
165	console.log(`
166	Total Experiences: ${metrics.totalExperiences}
167	Patterns Learned: ${metrics.patternsLearned}
168	Strategy Success Rate: ${metrics.strategySuccessRate}
169	Improvement Over Time: ${metrics.improvement}
170	`);
171	```
172
173	## Best Practices
174
175	1. Record consistently: Log all task outcomes, not just successes
176	2. Provide context: Rich context improves pattern matching
177	3. Set thresholds: Filter low-confidence learnings
178	4. Review periodically: Audit learned patterns for quality
179	5. Use vector search: Enable semantic pattern matching
180
181	## Troubleshooting
182
183	### Issue: Poor recommendations
184	Solution: Ensure sufficient training data (100+ experiences per task type)
185
186	### Issue: Slow pattern matching
187	Solution: Enable vector indexing in AgentDB
188
189	### Issue: Memory growing large
190	Solution: Set TTL for old experiences or enable pruning
191
192	## Learn More
193
194	- ReasoningBank Guide: agentic-flow/src/reasoningbank/README.md
195	- AgentDB Integration: packages/agentdb/docs/reasoningbank.md
196	- Pattern Learning: docs/reasoning/patterns.md

People who liked this also viewed

6 picks

Type

reasoningbank-intelligence

301stars24forks

View on GitHub

How to install reasoningbank-intelligence?

spencermarx/open-code-review/reasoningbank-intelligence

$npx -y skills add spencermarx/open-code-review --skill reasoningbank-intelligence

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# ReasoningBank Intelligence
2
3	## What This Skill Does
4
5	Implements ReasoningBank's adaptive learning system for AI agents to learn from experience, recognize patterns, and optimize strategies over time. Enables meta-cognitive capabilities and continuous improvement.
6
7	## Prerequisites
8
9	- agentic-flow v3.0.0-alpha.1+
10	- AgentDB v3.0.0-alpha.10+ (for persistence)
11	- Node.js 18+
12
13	## Quick Start
14
15	```typescript
16	import { ReasoningBank } from 'agentic-flow/reasoningbank';
17
18	// Initialize ReasoningBank
19	const rb = new ReasoningBank({
20	persist: true,
21	learningRate: 0.1,
22	adapter: 'agentdb' // Use AgentDB for storage
23	});
24
25	// Record task outcome
26	await rb.recordExperience({
27	task: 'code_review',
28	approach: 'static_analysis_first',
29	outcome: {
30	success: true,
31	metrics: {
32	bugs_found: 5,
33	time_taken: 120,
34	false_positives: 1
35	}
36	},
37	context: {
38	language: 'typescript',
39	complexity: 'medium'
40	}
41	});
42
43	// Get optimal strategy
44	const strategy = await rb.recommendStrategy('code_review', {
45	language: 'typescript',
46	complexity: 'high'
47	});
48	```
49
50	## Core Features
51
52	### 1. Pattern Recognition
53	```typescript
54	// Learn patterns from data
55	await rb.learnPattern({
56	pattern: 'api_errors_increase_after_deploy',
57	triggers: ['deployment', 'traffic_spike'],
58	actions: ['rollback', 'scale_up'],
59	confidence: 0.85
60	});
61
62	// Match patterns
63	const matches = await rb.matchPatterns(currentSituation);
64	```
65
66	### 2. Strategy Optimization
67	```typescript
68	// Compare strategies
69	const comparison = await rb.compareStrategies('bug_fixing', [
70	'tdd_approach',
71	'debug_first',
72	'reproduce_then_fix'
73	]);
74
75	// Get best strategy
76	const best = comparison.strategies[0];
77	console.log(`Best: ${best.name} (score: ${best.score})`);
78	```
79
80	### 3. Continuous Learning
81	```typescript
82	// Enable auto-learning from all tasks
83	await rb.enableAutoLearning({
84	threshold: 0.7, // Only learn from high-confidence outcomes
85	updateFrequency: 100 // Update models every 100 experiences
86	});
87	```
88
89	## Advanced Usage
90
91	### Meta-Learning
92	```typescript
93	// Learn about learning
94	await rb.metaLearn({
95	observation: 'parallel_execution_faster_for_independent_tasks',
96	confidence: 0.95,
97	applicability: {
98	task_types: ['batch_processing', 'data_transformation'],
99	conditions: ['tasks_independent', 'io_bound']
100	}
101	});
102	```
103
104	### Transfer Learning
105	```typescript
106	// Apply knowledge from one domain to another
107	await rb.transferKnowledge({
108	from: 'code_review_javascript',
109	to: 'code_review_typescript',
110	similarity: 0.8
111	});
112	```
113
114	### Adaptive Agents
115	```typescript
116	// Create self-improving agent
117	class AdaptiveAgent {
118	async execute(task: Task) {
119	// Get optimal strategy
120	const strategy = await rb.recommendStrategy(task.type, task.context);
121
122	// Execute with strategy
123	const result = await this.executeWithStrategy(task, strategy);
124
125	// Learn from outcome
126	await rb.recordExperience({
127	task: task.type,
128	approach: strategy.name,
129	outcome: result,
130	context: task.context
131	});
132
133	return result;
134	}
135	}
136	```
137
138	## Integration with AgentDB
139
140	```typescript
141	// Persist ReasoningBank data
142	await rb.configure({
143	storage: {
144	type: 'agentdb',
145	options: {
146	database: './reasoning-bank.db',
147	enableVectorSearch: true
148	}
149	}
150	});
151
152	// Query learned patterns
153	const patterns = await rb.query({
154	category: 'optimization',
155	minConfidence: 0.8,
156	timeRange: { last: '30d' }
157	});
158	```
159
160	## Performance Metrics
161
162	```typescript
163	// Track learning effectiveness
164	const metrics = await rb.getMetrics();
165	console.log(`
166	Total Experiences: ${metrics.totalExperiences}
167	Patterns Learned: ${metrics.patternsLearned}
168	Strategy Success Rate: ${metrics.strategySuccessRate}
169	Improvement Over Time: ${metrics.improvement}
170	`);
171	```
172
173	## Best Practices
174
175	1. Record consistently: Log all task outcomes, not just successes
176	2. Provide context: Rich context improves pattern matching
177	3. Set thresholds: Filter low-confidence learnings
178	4. Review periodically: Audit learned patterns for quality
179	5. Use vector search: Enable semantic pattern matching
180
181	## Troubleshooting
182
183	### Issue: Poor recommendations
184	Solution: Ensure sufficient training data (100+ experiences per task type)
185
186	### Issue: Slow pattern matching
187	Solution: Enable vector indexing in AgentDB
188
189	### Issue: Memory growing large
190	Solution: Set TTL for old experiences or enable pruning
191
192	## Learn More
193
194	- ReasoningBank Guide: agentic-flow/src/reasoningbank/README.md
195	- AgentDB Integration: packages/agentdb/docs/reasoningbank.md
196	- Pattern Learning: docs/reasoning/patterns.md

People who liked this also viewed

6 picks

Type