Ollama Memory

PassAudited by ClawScan on May 10, 2026.

Overview

This is a coherent local memory skill, but it stores and reuses persistent user/agent context and references helper scripts that were not included for review.

This skill appears benign and purpose-aligned for local memory. Before using it, inspect any referenced local Python scripts, avoid storing passwords or secrets, and periodically review USER.md, MEMORY.md, SOUL.md, and daily memory files because they can influence future assistant behavior.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Information saved in these files may influence later assistant behavior and may include sensitive personal preferences or context.

Why it was flagged

The skill intentionally reuses persistent local files containing user information, assistant persona, and prior session memory across future sessions.

Skill content
每次新会话开始时:

1. 读取 `SOUL.md` — AI 人格
2. 读取 `USER.md` — 用户信息
3. 读取 `memory/YYYY-MM-DD.md` (今天 + 昨天)
4. **主会话**:读取 `MEMORY.md`
Recommendation

Review and edit the memory files regularly, avoid saving secrets, and only enable this skill if you want persistent local memory.

What this means

Incorrect or unwanted stored memories could persist and affect future assistant responses.

Why it was flagged

The skill directs updates to persistent memory and persona/user files. This is purpose-aligned, but these files can shape future context if inaccurate or poisoned.

Skill content
### 更新规则

### 必须更新的文件
1. **memory/YYYY-MM-DD.md** - 每次会话
2. **MEMORY.md** - 新教训/重要信息
3. **SOUL.md** - 人格变化
4. **USER.md** - 用户偏好变化
Recommendation

Require explicit user confirmation for durable memory updates, especially changes to USER.md, MEMORY.md, SOUL.md, or behavior-guiding files.

What this means

The security of the actual memory operations depends on local scripts that were not included in this review.

Why it was flagged

The skill references local Python helper scripts for init/add/search, but the provided artifact set contains only SKILL.md and no script contents for review.

Skill content
python3 ~/.openclaw/workspace/scripts/memory-system/context-memory.py init
Recommendation

Inspect the referenced Python scripts before running them, and ensure they only read/write the intended local memory database and files.

What this means

Installation and model download trust depends on the package manager and Ollama model source.

Why it was flagged

The setup requires installing Ollama and downloading an embedding model. This is expected for the stated local embedding purpose, but it depends on external package/model sources.

Skill content
brew install ollama

# 启动服务
ollama serve

# 下载 embedding 模型
ollama pull nomic-embed-text
Recommendation

Install Ollama from trusted sources and verify the model name/source before use.