← 返回 Skills 市场
280
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install bobo-context-cleanup
功能描述
分析并归档 OpenClaw workspace 中低价值 memory 记录,缓解上下文膨胀与启动变慢。适用于上下文冗余、维护整理场景。默认先分析与计划,执行归档前要求确认。
使用说明 (SKILL.md)
Context Cleanup
用于整理 workspace 的 memory 日志,降低冗余上下文负担。
关键资源
- 脚本:
cleanup.sh - 策略:
references/policy.md
标准流程(必须)
- 分析现状
./skills/context-cleanup/cleanup.sh analyze
- 生成计划(不执行)
./skills/context-cleanup/cleanup.sh plan
- 用户确认后执行归档
./skills/context-cleanup/cleanup.sh archive
可选参数
# 指定截止日期(早于该日期的记录可归档)
./skills/context-cleanup/cleanup.sh archive 2026-03-01
# 仅预览,不执行
./skills/context-cleanup/cleanup.sh archive --dry-run
# 非交互执行(需谨慎)
./skills/context-cleanup/cleanup.sh archive --yes
# 机器可读输出
./skills/context-cleanup/cleanup.sh analyze --json
./skills/context-cleanup/cleanup.sh plan --json
执行规则
- 默认归档,不做永久删除
- 先
plan后archive - 归档前必须获得用户确认(除非用户明确同意
--yes) - 不处理
MEMORY.md、specs/、AGENTS.md
输出模板
🧹 上下文清理计划
- Memory 文件:X
- 低价值候选:A
- 归档候选:B
是否按计划执行归档?
发布前自检
./skills/context-cleanup/cleanup.sh analyze
./skills/context-cleanup/cleanup.sh plan
./skills/context-cleanup/cleanup.sh archive --dry-run
安全使用建议
This skill appears to do exactly what it says: analyze and archive low-value Markdown memory files in the local workspace. Before installing or running it: 1) Note the skill is deprecated — prefer the newer 'context-cleanup' slug from the same owner. 2) Ensure you have a recent backup of your workspace or run archive with --dry-run first to verify candidates. 3) Install or ensure node is available if you plan to use --json output (the script calls node but node is not declared in metadata). 4) Be careful with non-interactive runs (--yes); test interactively first. 5) Confirm WORKSPACE resolves to the expected directory to avoid moving files from an unintended location.
功能分析
Type: OpenClaw Skill
Name: bobo-context-cleanup
Version: 0.2.1
The skill bundle is a utility designed to manage and archive workspace memory logs to reduce context clutter. The primary script, `cleanup.sh`, identifies low-value or aged markdown files and moves them to an archive directory, following a structured analyze-plan-archive workflow. While the script contains a minor implementation flaw in how it uses `node -e` to format JSON output (potential syntax errors or shell expansion issues), there is no evidence of malicious intent, data exfiltration, or unauthorized system access.
能力评估
Purpose & Capability
Name/description, SKILL.md, and cleanup.sh are consistent: they analyze, plan, and move low-value Markdown memory files into a local archive directory. The script operates only on workspace paths (memory/, archive/, AGENTS.md, specs/) and does not request network access or unrelated credentials. Minor mismatch: registry metadata lists no required binaries, SKILL.md metadata mentions bash, but the script also invokes node for JSON output — node is not declared as a requirement.
Instruction Scope
SKILL.md instructs the agent to run ./cleanup.sh analyze|plan|archive. The script's actions match these steps: collect files, classify low-value/old files, and (after confirmation) move candidates to memory/archive. It does modify workspace files (moves to archive), which is expected for this purpose. It does not read or transmit data to external endpoints. Note: the --yes flag allows non-interactive archival which could be risky if invoked unintentionally.
Install Mechanism
No install spec is provided (instruction + included shell script). Nothing is downloaded or extracted. This minimizes install-time risk.
Credentials
No credentials or config paths are requested. The script optionally uses WORKSPACE (with a sensible default) so environment access is proportional. However, the script requires bash (declared in SKILL.md) and also calls node for JSON outputs but node is not declared; the missing declared dependency could cause failures or unexpected behavior when JSON output is requested.
Persistence & Privilege
always is false and the skill is user-invocable. It does not request persistent elevated privileges or modify other skills' configs. It does perform filesystem writes within the workspace (moving files into archive), which is within its stated scope.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install bobo-context-cleanup - 安装完成后,直接呼叫该 Skill 的名称或使用
/bobo-context-cleanup触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v0.2.1
Deprecated: please migrate to context-cleanup. This slug is kept for backward compatibility only.
v0.2.0
首发 0.2:重写清理脚本,支持 --dry-run/--yes/--json,新增策略文档,统一中文说明,增加 metadata.version/updatedAt。
元数据
常见问题
DEPRECATED - Bobo Context Cleanup 是什么?
分析并归档 OpenClaw workspace 中低价值 memory 记录,缓解上下文膨胀与启动变慢。适用于上下文冗余、维护整理场景。默认先分析与计划,执行归档前要求确认。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 280 次。
如何安装 DEPRECATED - Bobo Context Cleanup?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install bobo-context-cleanup」即可一键安装,无需额外配置。
DEPRECATED - Bobo Context Cleanup 是免费的吗?
是的,DEPRECATED - Bobo Context Cleanup 完全免费(开源免费),可自由下载、安装和使用。
DEPRECATED - Bobo Context Cleanup 支持哪些平台?
DEPRECATED - Bobo Context Cleanup 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 DEPRECATED - Bobo Context Cleanup?
由 irideas(@irideas)开发并维护,当前版本 v0.2.1。
推荐 Skills