Memory Cache

High-performance temporary storage system using Redis. Supports namespaced keys (mema:*), TTL management, and JSON serialization for session context and API...

Security Scan

The skill's code and runtime instructions match its Redis caching description and requests only a Redis connection and python; minor metadata inconsistencies and operational cautions remain but the package is internally coherent.

Purpose & Capability

Name/description (Redis-backed cache with mema: namespace) aligns with the included Python script and shell helper. The functionality (set/get/scan/ttl/expire/ping) and declared dependencies (redis, python-dotenv, python3) are appropriate for the stated purpose.

Instruction Scope

SKILL.md instructs using a .env (env.example.txt) and running the provided cache_manager.py via python3 or scripts/cache.sh which creates a local virtualenv and installs requirements. The script loads environment variables (.env and the process environment) and only communicates with Redis; it does not contact external endpoints beyond the Redis server. Note: example command references $WORKSPACE path — runtime must ensure correct path mapping.

Install Mechanism

No remote downloads or arbitrary URLs; installation is local pip install -r requirements.txt performed by the provided script or by the SKILL.md metadata. Requirements are standard (redis, python-dotenv). The script will create a .venv directory inside the skill tree to install dependencies.

Credentials

The runtime requires REDIS_URL (and supports REDIS_HOST/PORT/PASSWORD/DB/timeouts), which is proportional to a Redis cache skill. However, top-level registry metadata lists no required env vars while SKILL.md metadata declares REDIS_URL — this mismatch should be clarified. The script reads .env and environment variables, so any secrets present in .env will be loaded.

Persistence & Privilege

always is false and model invocation is allowed (default), which is appropriate. The skill writes a local .venv but does not modify other skills or system-wide agent configurations. No elevated privileges or permanent always-on presence are requested.

Assessment

This skill appears to do what it says: a Redis-backed cache accessed via a local Python script. Before installing, confirm: (1) REDIS_URL points to a trusted Redis instance (a misconfigured or public Redis can leak or accept data); (2) you understand that the skill will read a .env file and environment variables (do not store unrelated secrets there); (3) the script will create a .venv inside the skill directory and install packages from requirements.txt (no external downloads); (4) keys can be up to 512 MiB in value — avoid storing sensitive or large blobs unless intended. Also ask the publisher to fix the metadata mismatch (registry shows no required env vars while SKILL.md requires REDIS_URL) and to confirm the intended workspace path usage ($WORKSPACE). If you need tighter controls, restrict network access to the Redis host and avoid using the cache for secrets or long-term storage.

Standardized Redis-backed caching system for OpenClaw agents.

Prerequisites

• Binary: python3 must be available on the host.
• Credentials: REDIS_URL environment variable (e.g., redis://localhost:6379/0).

Setup

1. Copy env.example.txt to .env.
2. Configure your connection in .env.
3. Dependencies are listed in requirements.txt.

Core Workflows

1. Store and Retrieve

• Store: python3 $WORKSPACE/skills/memory-cache/scripts/cache_manager.py set mema:cache:<name> <value> [--ttl 3600]
• Fetch: python3 $WORKSPACE/skills/memory-cache/scripts/cache_manager.py get mema:cache:<name>

2. Search & Maintenance

• Scan: python3 $WORKSPACE/skills/memory-cache/scripts/cache_manager.py scan [pattern]
• Ping: python3 $WORKSPACE/skills/memory-cache/scripts/cache_manager.py ping

Key Naming Convention

Strictly enforce the mema: prefix:

• mema:context:* – Session state.
• mema:cache:* – Volatile data.
• mema:state:* – Persistent state.