Neural Pattern Match

2. Load the Pattern Index

Read memory/patterns/index.json. For each pattern entry, note:

• id, name, tags, confidence, last_used

3. Score Each Pattern

For every pattern in the index, calculate a relevance score using four weighted factors:

Relevance score = (tag_overlap * 0.4) + (recency * 0.2) + (confidence * 0.3) + (usage * 0.1)

4. Rank and Filter

Sort patterns by relevance score descending. Apply filters:

• Exclude patterns with confidence < 0.2 (unreliable)
• Exclude patterns with relevance score < 0.15 (not relevant)
• Take the top 5 results

5. Load Full Pattern Details

For each of the top-5 matches, read the full pattern JSON from memory/patterns/pat-[slug].json. Extract the approach, tools_used, and agents_involved fields.

6. Produce Context Enrichment

Format the matches for consumption by the requesting agent:

## Relevant Patterns Found

### 1. [Pattern Name] — Score: [x.xx]
- **Approach**: [summary]
- **Tools**: [list]
- **Confidence**: [x.xx] (used [N] times, [N] successes)
- **Last used**: [date]

### 2. [Pattern Name] — Score: [x.xx]
...

7. Pass to Requesting Agent

Inject the context enrichment block into the agent's prompt or task description. The agent can choose to follow a matched pattern or ignore it if the current task differs.

Output

• Top-5 ranked pattern matches with scores and approach summaries
• Context enrichment block ready for agent consumption

Review Checklist

• Task keywords were extracted accurately from the current task
• All four scoring factors were applied with correct weights
• Low-confidence patterns (< 0.2) were excluded
• Full pattern details were loaded for top-5 matches only (not all patterns)
• Output is formatted for easy consumption by the requesting agent