Skip to content
Back to skills

Smart Code Memory

ASecurity

Fast working memory and targeted code symbol map

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 16, 2026
ai-agentspythonbash

Works with

  • cli

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned October 16, 2026

npx -y skills add Osama-Alzahrani-UQU/OmniAgentic --skill smart-code-memory --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Smart Code Memory?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Smart Code Memory
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/osama-alzahrani-uqu-smart-code-memory/badge)](https://www.skillsdirectory.com/skills/osama-alzahrani-uqu-smart-code-memory)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: smart-code-memory
description: Fast working memory and targeted code symbol map
---

# Smart Code Memory & Fast Symbol Retrieval (`smart-code-memory`)

The **Smart Code Memory** skill equips AI models with a fast, lightweight working memory index. It eliminates the fatal flaw of forgetting context across prompts, which forces models to repeatedly re-read entire multi-thousand-line files from start to finish, causing hallucination, token exhaustion, and context degradation.

---

## 1. Core Operating Principles

1. **Index-First Targeted Navigation (البحث الموجه بدلاً من القراءة الكاملة المهدورة)**:
   - Before reading an entire file, query the local working memory index using `python scripts/code_memory.py lookup <path> <symbol>` or `outline <path>`.
   - The engine returns the exact file path, function/class name, signature, and precise line range (`start_line` to `end_line`).
   - Call `view_file` strictly on that targeted line range (`StartLine`, `EndLine`), saving 90%+ of tokens and preserving model focus.

2. **Explicit Fallback Protocol (المراجعة الشاملة عند عدم العثور على المشكلة)**:
   - **القاعدة الإلزامية**: إذا لم يجد النموذج سبب المشكلة أو العطل داخل النطاق المستهدف المحدد، أو إذا لم يكن الرمز مفهرساً:
     - **يُمنع التخمين أو التوقف**: ينتقل النموذج فوراً وتلقائياً لمراجعة الكود البرمجي بالكامل (`view_file` لكامل الملف أو السياق المحيط) للتأكد من تفاصيل التنفيذ والربط الشامل.

3. **Persistent Scratchpad Working Memory (الذاكرة المرجعية السريعة)**:
   - Store critical session facts, active bug root causes, target files, and user preferences in the lightweight memory scratchpad:
     ```bash
     python scripts/code_memory.py note set <key> "<val>"
     python scripts/code_memory.py note get <key>
     ```
   - On subsequent turns, query the scratchpad first instead of re-discovering previously analyzed logic.

---

## 2. CLI Tooling Reference (`scripts/code_memory.py`)

The companion engine runs as a global command (`code_memory` or `code-memory`) or direct Python script.

### 1. Indexing & Scanning
```bash
# Scan a directory or single file (generates .code_memory/index.json)
code_memory --root . scan

# Or run via local Python script:
# python scripts/code_memory.py --root . scan

# Force re-scan on file modification
code_memory --root . scan --force
```

### 2. Symbol Lookup
```bash
# Locate function, class, or method line numbers in milliseconds
python scripts/code_memory.py --root . lookup calculate_total

# Output as machine-readable JSON
python scripts/code_memory.py --root . lookup authenticate --json
```

### 3. File Outline & Structure Skeleton
```bash
# View complete outline of a file with line numbers
python scripts/code_memory.py --root . outline src/controllers/user.py
```

### 4. Working Memory Notes
```bash
# Save an active finding
python scripts/code_memory.py --root . note set bug_root_cause "Off-by-one error at lines 142-145"

# Retrieve an active finding
python scripts/code_memory.py --root . note get bug_root_cause

# List all stored notes
python scripts/code_memory.py --root . note list
```

---

## 3. Agent Execution Workflow

```
+-------------------------------------------------------------+
| User Request Received                                       |
+-------------------------------------------------------------+
                               |
                               v
+-------------------------------------------------------------+
| 1. Query Code Memory (`code_memory.py lookup <symbol>`)      |
+-------------------------------------------------------------+
          |                                       |
     [Found]                                 [Not Found]
          v                                       v
+----------------------------------+   +----------------------+
| 2. Read exact line range         |   | Fallback: Read full  |
| (`view_file(StartLine, EndLine)`)|   | file or grep keyword |
+----------------------------------+   +----------------------+
          |                                       |
          +-------------------+-------------------+
                              v
+-------------------------------------------------------------+
| 3. Problem resolved within range?                           |
|    - YES -> Apply surgical edit                             |
|    - NO  -> Fallback to reading full file & related modules |
+-------------------------------------------------------------+
                              v
+-------------------------------------------------------------+
| 4. Update memory index & note findings                      |
+-------------------------------------------------------------+
```

Files in this skill

  • SKILL.md4.9 KB
  • scripts/code_memory.py15.4 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.