القائمة
Store universal coding patterns into vector database. Auto-invokes after difficult tasks with broadly-applicable lessons. Trigger with "--store" or when user expresses frustration (strong learning signals).
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | coder-memory-store |
| description | Store universal coding patterns into vector database. Auto-invokes after difficult tasks with broadly-applicable lessons. Trigger with "--store" or when user expresses frustration (strong learning signals). |
The memory-only agent has ZERO access to Read/Write/Edit/Glob/Bash - it can ONLY use MCP memory tools. This prevents file reading pollution by design.
Store 0-1 insights per task (rarely 2-3).Store when: Non-obvious bugs, hard-won failure lessons, universal patterns across projects, user frustration signals.
Skip when: Standard practices, project-specific config, routine fixes, vague insights.
**Step 1: Format memory** ``` **Title:** [Concise title] **Description:** [2-3 sentence summary - CRITICAL for search]Content: [What happened, what was tried, what worked/failed, key lesson]
Tags: #role #topic #success|#failure
**Step 2: Extract metadata**
Parse the formatted text to extract:
- `title`: Plain text without markdown (from "**Title:**")
- `description`: Plain text without markdown (from "**Description:**")
- `tags`: Array of tags (e.g., ["#backend", "#jwt", "#auth"])
**Step 3: Detect role for storage**
Determine which role collection to store in based on task context. Use role_mapping below. Default to "universal" if unclear.
**Step 4: Search for duplicates**
Use `search_memory` with full formatted text as query, `roles=["detected_role", "universal"]`, `limit=10`.
Note: The `roles` parameter tells MCP which collections to search. Always include "universal" to catch cross-domain patterns.
**Step 5: Decide action**
- Near-identical exists → MERGE (combine, delete old)
- Related exists → UPDATE (enhance existing)
- Pattern emerges from 2+ episodic → GENERALIZE to semantic
- Different topic → CREATE new
**Step 6: Store**
Use `store_memory` with:
- `document`: Full formatted text (Title + Description + Content + Tags)
- `metadata`:
```json
{
"memory_type": "episodic|procedural|semantic",
"role": "backend|frontend|devops|ai|...",
"title": "Plain text title",
"description": "2-3 line summary",
"tags": ["#tag1", "#tag2"],
"confidence": "high|medium|low",
"frequency": 1
}
<role_mapping> Available roles (maps to Qdrant collections):
Default to "universal" if unclear. Each role maps to a collection named "{role}-patterns" (e.g., "backend-patterns"). </role_mapping>
<frustration_signals> User frustration = critical learning moment. When detected (profanity, "this is ridiculous", emotional language): Store as episodic with #failure #strong-signal tags. </frustration_signals>