← Back to Skills Marketplace
huyaohuahk

Task Memory

by huyaohuahk · GitHub ↗ · v1.2.0 · MIT-0
cross-platform ⚠ suspicious
114
Downloads
0
Stars
0
Active Installs
4
Versions
Install in OpenClaw
/install task-memory
Description
任务遗忘防护系统 — 解决 AI Agent 任务发出但未执行的记忆漏洞问题。 当需要创建、追踪、管理长期任务时使用,特别是:提出或承诺了某项计划后、设置 cron/自动化任务时、任务状态变更后、晨间/心跳检查时。 核心功能:通过 todo.json 持久化任务状态,todo_manager.py 管理增/查/改...
Usage Guidance
This skill appears to be a legitimate todo manager, but it's inconsistent: the documentation says todo.json will live in the skill/current directory (references/todo.json), while the script writes to /home/openclaw/.openclaw/workspace/backtest/todo.json. Before installing or running: 1) Do not run it as-is on a machine where /home/openclaw/... might exist and contain important data. 2) Inspect and edit scripts/todo_manager.py to point TODO_FILE to a safe, relative path (for example ./references/todo.json) or make the path configurable (env var or command-line option). 3) Run the script in a sandbox or test environment first to confirm behavior. 4) If you expect IM push integration, note the script does not implement it — that would need explicit, separate code. These mismatches are likely an oversight but are material; fix the file-path handling (or require/declare a configurable path) before trusting it with real data.
Capability Analysis
Type: OpenClaw Skill Name: task-memory Version: 1.2.0 The task-memory skill is a legitimate utility designed to help AI agents track long-term tasks and prevent 'memory loss' between sessions. The core logic in scripts/todo_manager.py is a straightforward Python script that performs CRUD operations on a local JSON file (todo.json) without any network activity, shell execution, or access to sensitive system files. The instructions in SKILL.md are well-aligned with the stated purpose of task persistence and do not contain any malicious prompt injections or unauthorized commands.
Capability Assessment
Purpose & Capability
SKILL.md describes a simple, local todo.json stored in the skill's references/ directory (or current directory). The shipped script instead uses a hardcoded TODO_FILE = "/home/openclaw/.openclaw/workspace/backtest/todo.json". This does not match the stated default path and indicates the code will read/write outside the skill bundle (and possibly overwrite or create files in an unexpected user home path).
Instruction Scope
SKILL.md instructs agents to run the script for add/check/list/done/etc and implies use of references/todo.json and optional push-to-IM in heartbeats. The script implements add/check/list/done operations but: (a) ignores the references/todo.json shipped in the package (because of the hardcoded path), and (b) does not itself implement network push to QQ/IM (it only prints to stdout). The runtime instructions therefore assume a different file layout/behavior than the code actually uses.
Install Mechanism
No install spec or external downloads; the skill is instruction+script only. No packages or remote installs are performed.
Credentials
The skill declares no required env vars or config paths, but the script requires filesystem write access and uses a hardcoded absolute path under /home/openclaw/.openclaw/... rather than a relative path or configurable location. That is disproportionate to the stated purpose (which implied local skill-scoped storage) and could unintentionally read/modify files in a host user's home.
Persistence & Privilege
The skill persists data to disk (normal for a todo tool). It does not request elevated platform privileges nor set always:true. The main issue is the persistence location (absolute path) which gives broader filesystem presence than the docs imply.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install task-memory
  3. After installation, invoke the skill by name or use /task-memory
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.2.0
v1.2.0: --add强制校验deadline格式,--done自动归档30天,新增--archive/--purge
v1.1.0
v1.1.0: SKILL.md去除所有个人信息,改为通用路径,CHANGELOG去除身份信息,适合所有用户部署
v1.0.1
v1.0.1: 增加 CHANGELOG.md 知识库,完整版本历史/设计决策/未来计划
v1.0.0
首发版:任务遗忘防护系统 todo_manager.py + todo.json,自动追踪承诺的计划
Metadata
Slug task-memory
Version 1.2.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 4
Frequently Asked Questions

What is Task Memory?

任务遗忘防护系统 — 解决 AI Agent 任务发出但未执行的记忆漏洞问题。 当需要创建、追踪、管理长期任务时使用,特别是:提出或承诺了某项计划后、设置 cron/自动化任务时、任务状态变更后、晨间/心跳检查时。 核心功能:通过 todo.json 持久化任务状态,todo_manager.py 管理增/查/改... It is an AI Agent Skill for Claude Code / OpenClaw, with 114 downloads so far.

How do I install Task Memory?

Run "/install task-memory" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Task Memory free?

Yes, Task Memory is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Task Memory support?

Task Memory is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Task Memory?

It is built and maintained by huyaohuahk (@huyaohuahk); the current version is v1.2.0.

💬 Comments