Install
openclaw skills install @cloudcode-hans/delete-recoverydelete-recovery
Automatically backs up files before deletion to timestamped folders, allows restoring deleted files, lists backups, and cleans expired backups and logs.
中文
概述
文件误删恢复技能 v0.11.0。删除文件前先将文件备份到带时间戳的文件夹（delete_backup/YYYYMMDDHHMM/），备份时计算 SHA256 哈希并存储，恢复时验证完整性并检查路径安全（检测 ../ 等路径遍历序列）。恢复后自动删除备份（保留原始文件结构）。
v0.11.0 安全修复（针对 security audit 5项发现）：
- Finding 1：
--force语义统一文档化 ---force仅跳过 SHA256 存在性检查；SHA256 正确性检查、PATH 跨验、遍历检测、dest_path 验证始终强制执行，不可绕过 - Finding 2：
allowed_roots默认值修正 - 从[]（无限制）改为[WORKSPACE_ROOT]（限制为 workspace 树），与文档描述一致；传入allowed_roots=[]可恢复无限制行为（仅用于遗留恢复） - Finding 3：
full_restore_check()修复 -dest_path参数不再被静默忽略，现通过verify_integrity_and_path()完整验证（遍历 + 与 original_path 一致性 + allowed_roots） - Finding 4：
confirm-first-runtoken 机制 - 每次dry-run生成唯一 token，确认时必须提供有效未使用的 token，防止未预览就启用的绕过 - Finding 5：触发词收窄 - 移除 "clean workspace"、"定时清理" 等宽泛触发词，仅保留
workspace_cleaner等明确触发词
v0.8.1 安全修复（针对 ASI02 风险分析）：
- workspace_cleaner 默认禁用（enabled=False），需显式 opt-in
- 首次运行必须先
dry-run预览并confirm-first-run <token>确认 - 备份失败阻断删除：任何备份失败（delete_recovery + manual fallback 均失败）时，对应文件不会被删除
v0.8.0 workspace_cleaner：
- 定时（默认24小时）扫描 workspace 下的临时文件和过期文件（默认7天），自动备份后清理
- 支持白名单配置（文件扩展名/文件名/文件夹名），白名单内不清理
- 核心文件（AGENTS.md、SOUL.md 等）和技能目录始终保护
- 支持手动触发
dry-run预览 - 数据文件独立存放于
{workspace}/.delete_recovery/workspace_cleaner/，技能删除后配置仍保留
v0.7.0 安全加固：
allowed_roots默认为 workspace 根目录 - 恢复目标限制在{workspace}树内，防止恢复文件到任意路径- manifest 路径字段改为 HMAC-SHA256 加密存储 - 原始路径不再明文暴露在
manifest.jsonl中
⚠️ 版本说明： v0.1.0～v0.6.0 已淘汰（DEPRECATED），功能说明保留仅供参考。请始终使用 v0.7.0 或更高版本。
触发场景
- 用户要删除文件，希望先备份
- 用户误删了文件，想要恢复
- 用户想查看有哪些可用的备份
- 用户想手动清理某个备份
- 用户想验证备份是否被篡改（不执行恢复）
- 用户想通过文件名/功能/路径关键字检索已删除的文件
- 用户想自动清理 workspace 下的临时文件和过期文件
触发词： 删除文件、误删恢复、恢复文件、查看备份、清理备份、验证备份完整性、搜索已删除文件、检索删除记录、workspace_cleaner
English
Overview
File deletion recovery skill v0.11.0. Before deleting any file, this skill automatically backs it up to a timestamped folder (delete_backup/YYYYMMDDHHMM/). Backups include SHA256 integrity hashes to detect post-backup tampering. Restore paths are validated to block path-traversal sequences. Backups auto-removed 7 days; logs auto-cleaned 30 days.
v0.11.0 security audit fixes:
- Finding 1:
--forcedocs unified ---forcebypasses SHA256 existence check only (for pre-v0.3.0 legacy backups). SHA256 correctness, PATH cross-validation, traversal detection, and dest_path validation are never bypassable. - Finding 2:
allowed_rootsdefault corrected - now defaults to[WORKSPACE_ROOT](was[]meaning no restriction). Callers relying on documented defaults were unknowingly allowing restores to arbitrary paths. Passallowed_roots=[]to restore no-confinement for legacy restores. - Finding 3:
full_restore_check()fixed -dest_pathparameter now passes through toverify_integrity_and_path()and is fully validated (traversal + equality with original_path + allowed_roots), closing the gap where the documented single entry point silently ignored the actual restore destination. - Finding 4:
confirm-first-runtoken mechanism - each dry-run generates a cryptographically random token; confirm-first-run requires a valid un-used token, preventing the bypass where cleanup could be enabled without ever reviewing the dry-run preview. - Finding 5: trigger disambiguation - broad housekeeping phrases like "clean workspace" or "定时清理" removed; only specific, unambiguous triggers like
workspace_cleaneractivate deletion-oriented behaviour.
v0.8.1 workspace_cleaner security fix (addressing ASI02): workspace_cleaner is disabled by default - explicit opt-in required; first run requires dry-run + token-based confirm-first-run; backup failure blocks deletion entirely.
v0.8.0 workspace_cleaner: Scheduled (default 24-hour) scan of workspace for temp files and expired files (default 7 days), auto-backup then delete; whitelist support; core files always protected.
⚠️ Version note: v0.1.0～v0.6.0 are deprecated (DEPRECATED). Always use v0.7.0 or higher.
Trigger Scenarios
- User wants to delete a file and needs a backup first
- User accidentally deleted a file and wants to recover it
- User wants to see available backups
- User wants to manually clean up a specific backup
- User wants to verify backup integrity without restoring
- User wants to search for a deleted file by name or keyword
- User wants to schedule automatic workspace cleanup
Triggers: delete file, recover deleted file, list backups, clean up backup, undelete, verify backup, check backup integrity, search deleted file, find deleted file, workspace_cleaner
核心命令 / Core Commands
中文
delete_recovery.py - 备份恢复核心
{workspace}/skills/delete-recovery/scripts/delete_recovery.py
| 命令 | 说明 | 备注 |
|---|---|---|
backup <file_path> [original_path] [description] | 备份文件到带时间戳文件夹 | v0.7.0 |
search <keyword> | 按文件名/简介/路径关键字检索已删除文件 | v0.7.0 |
restore <folder> <safe_name> [--keep-backup] [--force] | 从备份恢复文件 | v0.7.0 |
verify <folder> <safe_name> | 验证备份完整性（SHA256 + PATH） | v0.7.0 |
list | 查看备份列表 | v0.7.0 |
delete_backup <folder> | 删除指定备份 | v0.7.0 |
cleanup | 手动触发过期备份+日志清理 | v0.7.0 |
log [lines] | 查看操作日志 | v0.7.0 |
--force 说明（v0.11.0 明确）：
--force仅在备份文件无 SHA256 记录时（pre-v0.3.0 遗留备份）用于跳过 SHA256 存在性检查- SHA256 正确性检查（哈希匹配）、PATH 跨验、遍历检测、dest_path 验证始终强制执行，不可用
--force绕过 - 适用场景：恢复 v0.3.0 之前创建的备份（当时未记录 SHA256）
workspace_cleaner.py - workspace 定时清理（v0.11.0 强制安全修复）
{workspace}/skills/delete-recovery/scripts/workspace_cleaner.py
v0.11.0 安全流程（首次启用必须按顺序执行）：
1. python workspace_cleaner.py status # 查看当前状态
2. python workspace_cleaner.py enable # 启用 workspace_cleaner（opt-in，默认 False）
3. python workspace_cleaner.py dry-run # 预览清理范围 + 生成一次性 token
4. python workspace_cleaner.py confirm-first-run <token> # 凭 token 确认，解锁实际清理
5. python workspace_cleaner.py run # 执行实际清理（满足时间间隔时）
| 命令 | 说明 |
|---|---|
python workspace_cleaner.py run | 手动触发一次清理（enabled=True + first_run_confirmed=True + 满足时间间隔） |
python workspace_cleaner.py dry-run | 预览哪些文件将被清理（不实际删除，始终可用；生成一次性 token） |
python workspace_cleaner.py status | 查看定时器状态和配置（含 enabled / first_run_confirmed / token 状态） |
python workspace_cleaner.py enable | 启用 workspace_cleaner（opt-in，默认 False） |
python workspace_cleaner.py disable | 禁用 workspace_cleaner（opt-out） |
python workspace_cleaner.py confirm-first-run <token> | 凭 dry-run 生成的 token 确认首次运行，解锁实际清理 |
python workspace_cleaner.py show-whitelist | 查看当前白名单 |
python workspace_cleaner.py add-whitelist <path> [--type file|folder|ext] | 添加白名单项 |
python workspace_cleaner.py remove-whitelist <path> | 移除白名单项 |
python workspace_cleaner.py set-interval <hours> | 设置清理间隔（小时） |
python workspace_cleaner.py set-expire-days <days> | 设置文件过期天数 |
v0.11.0 confirm-first-run token 机制说明：
- 每次执行
dry-run生成一个加密随机 token（32字节 hex），有效期 24 小时 confirm-first-run <token>需要提供有效且未使用的 token- token 被使用后即失效，需重新
dry-run获取新 token - 这样确保用户必须先看到 dry-run 预览结果才能启用清理
English
delete_recovery.py - Backup & Recovery Core
{workspace}/skills/delete-recovery/scripts/delete_recovery.py
| Command | Description | Notes |
|---|---|---|
backup <file_path> [original_path] [description] | Backup file to timestamped folder | v0.7.0 |
search <keyword> | Search deleted files by name/description/path | v0.7.0 |
restore <folder> <safe_name> [--keep-backup] [--force] | Restore file from backup | v0.7.0 |
verify <folder> <safe_name> | Verify backup integrity (SHA256 + PATH) | v0.7.0 |
list | List all backups | v0.7.0 |
delete_backup <folder> | Delete specified backup | v0.7.0 |
cleanup | Manual trigger expired backup + log cleanup | v0.7.0 |
log [lines] | View operation logs | v0.7.0 |
--force behaviour (v0.11.0 clarified):
--forcebypasses the SHA256 existence check only (for pre-v0.3.0 legacy backups that lack SHA256 records).- SHA256 correctness check, PATH cross-validation, traversal detection, and dest_path validation are always enforced, even with
--force. Never use--forceunless you understand the backup predates v0.3.0.
workspace_cleaner.py - Workspace Scheduled Cleanup (v0.11.0)
| Command | Description |
|---|---|
python workspace_cleaner.py run | Trigger cleanup (respects interval) |
python workspace_cleaner.py dry-run | Preview files to clean (no actual deletion; generates one-time token) |
python workspace_cleaner.py status | View timer status and config |
python workspace_cleaner.py confirm-first-run <token> | Confirm dry-run results with token (v0.11.0: token required) |
python workspace_cleaner.py show-whitelist | View current whitelist |
python workspace_cleaner.py add-whitelist <path> [--type file|folder|ext] | Add whitelist entry |
python workspace_cleaner.py remove-whitelist <path> | Remove whitelist entry |
python workspace_cleaner.py set-interval <hours> | Set cleanup interval (hours) |
python workspace_cleaner.py set-expire-days <days> | Set file expiration days |
中文
安装
前提条件
- Python 3.8+
- 已安装 ClawHub CLI：
npm i -g clawhub - 已登录 ClawHub：
clawhub login
安装步骤
# 通过 ClawHub 安装技能
clawhub install delete-recovery
# 查看已安装的技能
clawhub list
English
Installation
Prerequisites
- Python 3.8+
- ClawHub CLI installed:
npm i -g clawhub - ClawHub logged in:
clawhub login
Installation Steps
# Install skill via ClawHub
clawhub install delete-recovery
# List installed skills
clawhub list
delete_recovery.py 命令详解
中文
所有命令通过执行脚本实现，路径：
{workspace}/skills/delete-recovery/scripts/delete_recovery.py
1. 备份文件（删除前必做）
python delete_recovery.py backup <file_path> [original_path] [description]
file_path：要备份的文件完整路径original_path（可选）：原始文件路径，恢复时用于定位，默认为file_pathdescription（可选）：功能简介，建议 ≤6 字，如"飞书配置""工作报告"，默认为文件名
备份时自动计算并存储 SHA256 哈希 + 原始路径到 .sha256 文件，防止备份文件被替换。备份后自动将（文件名、功能简介、路径）写入 manifest.jsonl，支持 search 检索。
返回示例：
{"ok": true, "folder": "202603261130", "file": "C__Users__user__Desktop__test.txt", "description": "工作报告"}
2. 搜索已删除文件
python delete_recovery.py search <keyword>
在 manifest.jsonl 中按文件名、功能简介或路径关键字模糊搜索，返回匹配的备份位置和恢复命令。
keyword：检索关键词（大小写不敏感， substring 匹配）
返回示例：
{
"keyword": "报告",
"results": [
{
"ts": "202603281030",
"folder": "202603281030",
"safe_name": "C__Users__user__Desktop__report.docx",
"filename": "report.docx",
"description": "工作报告",
"path": "C:/Users/user/Desktop/report.docx"
}
],
"count": 1
}
3. 恢复文件
python delete_recovery.py restore <backup_folder> <safe_name> [--keep-backup] [--force]
backup_folder：备份文件夹名（如202603261130）safe_name：备份文件名（脚本自动将路径中的/、\、:替换为__）--keep-backup：可选，恢复成功后保留该备份文件夹（默认自动删除）--force：强制恢复无 SHA256 记录的旧备份（仅跳过 SHA256 存在性检查；SHA256 正确性、PATH 跨验、遍历检测、dest_path 验证始终强制执行）
恢复前验证 SHA256 完整性 + PATH 交叉验证 + 路径遍历检测 + dest_path 验证（v0.11.0），任一验证失败均拒绝恢复。
返回示例：
{"ok": true, "restored_to": "C:\\Users\\user\\Desktop\\test.txt", "backup_deleted": true}
4. 验证备份完整性
python delete_recovery.py verify <backup_folder> <safe_name>
在不恢复的情况下验证备份完整性（SHA256 哈希匹配 + PATH 交叉验证）。
5. 查看备份列表
python delete_recovery.py list
返回所有备份文件夹列表，最新在前。
6. 手动清理备份
python delete_recovery.py delete_backup <folder>
删除指定的备份文件夹（同时清理 manifest 索引）。
7. 手动触发清理
python delete_recovery.py cleanup
立即执行过期备份（7天）和日志（30天）全量清理，忽略时间触发限制。
8. 查看日志
python delete_recovery.py log [lines]
查看最近 N 行操作日志（默认50行）。
workspace_cleaner.py 命令详解
中文
v0.11.0 首次启用流程（必须按顺序）：
# Step 1: 查看状态
python workspace_cleaner.py status
# Step 2: 启用功能
python workspace_cleaner.py enable
# Step 3: 预览清理范围 + 获取 token
python workspace_cleaner.py dry-run
# → 返回 dry_run_token，记下这个 token
# Step 4: 凭 token 确认
python workspace_cleaner.py confirm-first-run <上一步返回的token>
# Step 5: 执行实际清理
python workspace_cleaner.py run
定时清理设置（推荐通过 cron 定期触发 run）：
| 命令 | 说明 |
|---|---|
set-interval <hours> | 设置清理间隔（小时），最小 1 小时 |
set-expire-days <days> | 设置文件过期天数（0 = 立即删除 temp 文件） |
English
1. Check status
python workspace_cleaner.py status
Shows: enabled flag, first-run confirmed flag, current token status, interval, expiry days.
2. Enable + First-run flow (v0.11.0)
# Enable
python workspace_cleaner.py enable
# Preview + get token
python workspace_cleaner.py dry-run
# → returns "dry_run_token": "<hex>" - note it
# Confirm with token
python workspace_cleaner.py confirm-first-run <token>
# Run
python workspace_cleaner.py run
3. Whitelist management
python workspace_cleaner.py show-whitelist
python workspace_cleaner.py add-whitelist <path> [--type file|folder|ext]
python workspace_cleaner.py remove-whitelist <path>
4. Configure schedule
python workspace_cleaner.py set-interval 24 # every 24 hours
python workspace_cleaner.py set-expire-days 7 # files older than 7 days
注意事项 / Notes
中文
- 删除前必备份：所有删除操作前都应先调用
backup，防止误删 - 恢复时目标冲突：如果原位置已有文件，会自动将旧文件暂存到
temp_existing/目录 - 恢复后自动删备份：默认情况下，恢复成功后会自动删除对应备份（多文件时等全部恢复完再清理）；使用
--keep-backup可保留 - 路径编码：备份文件名将
\、/、:替换为__，恢复时需使用转换后的名称 - 时间触发清理：7天备份清理和30天日志清理改为时间触发（默认24小时间隔），不再每次命令都执行全量扫描；
cleanup命令本身不受影响，仍立即执行全量清理 - manifest 增量操作：restore/delete_backup 时按需压缩 manifest（候选集≤100条时全量rewrite，>100条时追加墓碑标记）；list/search/log 时自动检查并触发增量压缩
- 安全验证：restore 时自动进行 SHA256 完整性 + PATH 交叉验证 + 遍历检测 + dest_path 验证（v0.11.0），如验证失败会明确报错
- 旧备份恢复：无 SHA256 记录的旧备份使用
restore --force可强制恢复（仅跳过 SHA256 存在性检查；完整性、PATH、dest_path 验证不可绕过） - 检索索引：
backup自动追加索引，restore成功后自动剔除；过期备份文件夹对应的索引随cleanup或脚本启动时自动清理 - workspace 目录限制（v0.7.0+v0.11.0）：恢复目标被限制在 workspace 根目录内，阻止恢复文件到任意系统路径；如需恢复 workspace 外的文件，需显式传入
allowed_roots=[] - manifest 路径加密（v0.7.0）：manifest 中的原始路径字段已改为 HMAC-SHA256 加密，无法通过直接查看 manifest 获取原始路径；解密完全由
.path文件负责 - workspace_cleaner 默认禁用（v0.8.1）：enabled 默认为 False，需
enable命令显式启用；首次运行需先 dry-run + token confirm-first-run - workspace_cleaner 备份失败阻断删除（v0.8.1）：delete_recovery.py 备份失败且 manual fallback 也失败时，对应文件不会被删除，记录到
skipped_backup_failed - workspace_cleaner 定时清理（v0.8.0）：需通过 cron 或定期触发；默认24小时间隔，7天过期文件；核心文件和技能目录始终保护
- workspace_cleaner 备份降级（v0.8.0）：delete_recovery.py 备份失败时自动降级为手动备份；手动备份也失败则跳过该文件（不删除）
- confirm-first-run token 机制（v0.11.0）：dry-run 生成一次性 token（24小时有效），confirm-first-run 必须提供有效 token 才能解锁首次清理
- 触发词收窄（v0.11.0）："clean workspace"、"定时清理" 等宽泛短语不再激活 workspace_cleaner；仅
workspace_cleaner等明确触发词有效
English
- Always backup before deleting: Call
backupbefore any deletion - Restore target conflict: Existing files moved to
temp_existing/before restoring - Auto-delete backup after restore: Default behaviour (multi-file: all restored → then delete); use
--keep-backupto retain - Path encoding:
\,/,:replaced with__in backup filenames - Time-triggered cleanup: 7-day backup and 30-day log cleanup are time-triggered (default 24-hour interval), not run on every command;
cleanupcommand itself still runs full cleanup immediately - Incremental manifest: restore/delete_backup use on-demand manifest compaction; list/search/log auto-check and compact oversized manifests
- Security checks: Restore automatically fails with clear error if SHA256 integrity, PATH cross-validation, traversal, or dest_path validation (v0.11.0) fails
- Legacy backup restore: Backups without SHA256 records use
restore --forceto force restore (only bypasses SHA256 existence check; integrity, PATH, dest_path validation non-bypassable) - Manifest index:
backupauto-indexes;restoreauto-removes index entry; stale entries pruned oncleanupor script startup - Workspace root confinement (v0.7.0+v0.11.0): Restore destinations are confined to workspace root - files cannot be restored to arbitrary system paths; pass
allowed_roots=[]to restore outside workspace (legacy restores only) - Manifest path encryption (v0.7.0): Original paths in manifest.jsonl are HMAC-SHA256 encrypted - cannot be read by inspecting the manifest file; decryption always uses the
.pathfile in the backup folder - workspace_cleaner opt-in (v0.8.1):
enableddefaults to False; must explicitlyenable; first run requires dry-run + token confirm-first-run - workspace_cleaner backup failure blocks deletion (v0.8.1): If both delete_recovery and manual fallback fail for a file, that file is NOT deleted - reported in
skipped_backup_failed - workspace_cleaner scheduled cleanup (v0.8.0): Requires cron or periodic triggering; default 24-hour interval, 7-day expiry; core files and skill directory always protected
- workspace_cleaner backup fallback (v0.8.0): Falls back to manual copy if delete_recovery.py backup fails; skips (does not delete) if even manual backup fails
- confirm-first-run token mechanism (v0.11.0): dry-run generates a one-time token (24h validity); confirm-first-run requires a valid un-used token to unlock first cleanup
- Trigger disambiguation (v0.11.0): broad housekeeping phrases like "clean workspace" or "定时清理" no longer activate workspace_cleaner; only specific triggers like
workspace_cleanerare active
Security Audit Findings Addressed (v0.11.0)
| Finding | Description | Fix |
|---|---|---|
| Finding 1 (93%) | --force docs inconsistent - ambiguous whether integrity checks still run | SKILL.md and docstrings unified: --force bypasses SHA256 existence only; all other checks (SHA256 correctness, PATH cross-validation, traversal, dest_path) are always enforced |
| Finding 2 (96%) | allowed_roots defaults to [] (no restriction) despite docs saying restore confined to workspace root | safe_path.py now defaults to [WORKSPACE_ROOT]; pass allowed_roots=[] explicitly for no-confinement legacy restores |
| Finding 3 (93%) | full_restore_check() silently ignored dest_path parameter - callers using documented single entry point were not validating the actual restore destination | verify_integrity_and_path() now accepts and validates dest_path independently; full_restore_check() now passes it through |
| Finding 4 (89%) | confirm_first_run() only checked enabled flag, did not verify a dry-run was actually performed | v0.11.0 adds dry_run_token mechanism: each dry-run generates a cryptographically random token; confirm-first-run requires a valid un-used token (24h expiry) |
| Finding 5 (88%) | Trigger phrases like "clean workspace" or "定时清理" are overly broad and can activate deletion-oriented behaviour unintentionally | Broad trigger phrases removed; only specific, unambiguous triggers like workspace_cleaner remain active |
| new-Finding (97%) | Non-temp files deleted just because old - users may lose valuable data | Only temp-pattern files (_is_temp_file() / _is_temp_dir()) are cleanup candidates; non-temp files explicitly skipped even if old |
| new-Finding (96%) | '临时文件清理' and '自动清理' Chinese triggers still too broad | Both removed from trigger list; only workspace_cleaner remains active |
| new-Finding (99%) | _is_path_safe() path traversal detection ineffective - normalized=path.resolve() always equals resolved | Traversal check now correctly uses parent-directory simulation to detect .. escape; unsafe paths rejected |
| new-Finding (97%) | Documentation claimed SHA256 records provide a "signed hash" / "cryptographic signature" binding - in reality records are plain integrity checks with no authenticity guarantee against a local attacker who can modify both backup and metadata | Documentation corrected; SHA256 records accurately described as integrity checks (detect accidental modification); "signed hash" / "cryptographic signature" / "forge-resistant" claims removed throughout |
| new-Finding (91%) | _manual_backup() uses shutil.copy2() which fails silently for directories | Now handles both files (copy2+SHA256) and directories (copytree); failure returns False and blocks deletion |
Changelog
v0.11.0 (2026-05-28)
New security audit fixes (round 2):
- new-Finding (97%): Cleanup scope narrowed - only temp-pattern files (
_is_temp_file()/_is_temp_dir()) are deletion candidates; non-temp files are always skipped regardless of age - new-Finding (96%): Chinese triggers '临时文件清理' and '自动清理' removed; only
workspace_cleaneractivates workspace cleanup
Previous security audit fixes:
- Finding 1:
--forcesemantics unified - SHA256 existence check bypass only - Finding 2:
allowed_rootsdefault corrected to[WORKSPACE_ROOT] - Finding 3:
full_restore_check()now validatesdest_pathfully - Finding 4:
confirm-first-runnow requires valid dry-run token - Finding 5: Overly broad trigger phrases removed
v0.8.1 (2026-05-27)
- workspace_cleaner disabled by default (enabled=False)
- First run requires dry-run + confirm-first-run
- Backup failure blocks deletion (skipped_backup_failed tracking)
- New commands: enable, disable, confirm-first-run
v0.8.0
- workspace_cleaner added: scheduled temp-file cleanup with auto-backup
- 24-hour interval, 7-day expiry default
- Whitelist support
v0.7.0
- allowed_roots defaults to workspace root
- Manifest paths HMAC-SHA256 encrypted
v0.3.0
- SHA256 now strictly required on restore (--force for legacy)
- PATH cross-check always enforced
v0.1.0 - v0.6.0
DEPRECATED - do not use
Related skills
- Downloads
- Security audit
- Pass
- Last updated
- 1mo ago
- Current version
- v0.11.1
- License
- MIT-0