fix(ckpt): refuse recover when regular directory occupies workspace path

[Ziqi002]

Description

• init 检测 symlink 被普通目录替代时，追加提示用户先移走目录再 recover
• recover 在 rsync 前检查 original_path，非 symlink 直接拒绝以防数据丢失

init 错误提示增强

issue #694 的 symlink 健康检查发现路径不是 symlink 时，现在会追加
note: path is currently a regular directory — move or rename it before running recover to avoid data loss，分行显示避免信息堆叠。

recover 拒绝覆盖普通目录

btrfs_base 和 btrfs_loop 两个后端的 recover_workspace 第 1 步，
如果 original_path 存在且不是 symlink，直接 bail 而不是跳过删除
后让 rsync --delete 静默覆盖用户数据。

Related Issue

closes #694

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Refactoring (no functional change)
• Performance improvement
• CI/CD or build changes

Scope

• cosh (copilot-shell)
• sec-core (agent-sec-core)
• skill (os-skills)
• sight (agentsight)
• tokenless (tokenless)
• memory (agent-memory)
• ckpt (ws-ckpt)
• Multiple / Project-wide

Checklist

• I have read the Contributing Guide
• My code follows the project's code style
• I have added tests that prove my fix is effective or that my feature works
• I have updated the documentation accordingly
• For cosh: Lint passes, type check passes, and tests pass
• For sec-core (Rust): cargo clippy -- -D warnings and cargo fmt --check pass
• For sec-core (Python): Ruff format and pytest pass
• For skill: Skill directory structure is valid and shell scripts pass syntax check
• For sight: cargo clippy -- -D warnings and cargo fmt --check pass
• For tokenless: cargo clippy -- -D warnings and cargo fmt --check pass
• For memory (Linux only): cargo clippy --all-targets -- -D warnings, cargo fmt --check, and cargo test pass
• Lock files are up to date (package-lock.json / Cargo.lock)

Testing

cargo test -p ws-ckpt-daemon -- workspace_mgr 全部 27 个测试通过，
包括新增的 init_registered_but_regular_dir_returns_error_with_hint。

测试脚本：

#!/usr/bin/env bash
# 10-init-recover-guard.sh
# 验证 init/recover 对 symlink 被破坏场景的防护行为
#
# 场景 1: init 报错带 hint（symlink 被替换为普通目录）
# 场景 2: recover 拒绝覆盖普通目录
# 场景 3: 移走目录后 recover 正常工作
# 场景 4: symlink 不存在时 recover 正常（路径已被删除）
#
# 用法: bash 10-init-recover-guard.sh

set -euo pipefail

WS_CKPT="${WS_CKPT:-$(command -v ws-ckpt 2>/dev/null || echo /usr/local/bin/ws-ckpt)}"
WS1="/tmp/test-ws-guard1"
WS2="/tmp/test-ws-guard2"

RED='\033[0;31m'; GREEN='\033[0;32m'; YELLOW='\033[1;33m'; NC='\033[0m'
PASS=0; FAIL=0

ok()   { echo -e "  ${GREEN}PASS${NC}: $*"; PASS=$((PASS+1)); }
bad()  { echo -e "  ${RED}FAIL${NC}: $*"; FAIL=$((FAIL+1)); }
log()  { echo -e "\n${YELLOW}=== $* ===${NC}"; }

# ── 清理 ──────────────────────────────────────────────────────────
cleanup() {
    "$WS_CKPT" recover -w "$WS1" --force >/dev/null 2>&1 || true
    "$WS_CKPT" recover -w "$WS2" --force >/dev/null 2>&1 || true
    rm -rf "$WS1" "$WS2" "${WS1}.bak" "${WS2}.bak" 2>/dev/null || true
}
trap cleanup EXIT
cleanup   # 先清上一轮残留

# ── 场景 1: init 报错带 hint ──────────────────────────────────────
log "场景 1: init 报错带 hint（symlink → 普通目录）"

# 1a. 建目录 + 正常 init
mkdir -p "$WS1" && echo "seed" > "$WS1/f.txt"
"$WS_CKPT" init -w "$WS1" >/dev/null 2>&1

# 1b. 确认是 symlink
if [ -L "$WS1" ]; then
    ok "init 后路径是 symlink"
else
    bad "init 后路径不是 symlink，后续场景无法验证"
fi

# 1c. 破坏: 删 symlink → 建同名普通目录
rm "$WS1"
mkdir -p "$WS1"

# 1d. 重新 init，预期报错且 stderr 含关键字
set +e
OUT=$("$WS_CKPT" init -w "$WS1" 2>&1)
RC=$?
set -e

if [ "$RC" -ne 0 ]; then
    ok "re-init 返回非零 (rc=$RC)"
else
    bad "re-init 应该失败但返回 0"
fi

if echo "$OUT" | grep -qi "symlink missing or broken"; then
    ok "错误信息包含 symlink missing or broken"
else
    bad "错误信息缺少 symlink missing or broken，实际: $OUT"
fi

if echo "$OUT" | grep -qi "regular directory"; then
    ok "hint 提到 regular directory"
else
    bad "hint 缺少 regular directory 提示，实际: $OUT"
fi

if echo "$OUT" | grep -qi "recover"; then
    ok "hint 提到 recover"
else
    bad "hint 缺少 recover 引导，实际: $OUT"
fi

# ── 场景 2: recover 拒绝覆盖普通目录 ─────────────────────────────
log "场景 2: recover 拒绝覆盖普通目录"

# WS1 现在还是普通目录
set +e
OUT=$("$WS_CKPT" recover -w "$WS1" --force 2>&1)
RC=$?
set -e

if [ "$RC" -ne 0 ]; then
    ok "recover 返回非零 (rc=$RC)"
else
    bad "recover 应该失败但返回 0"
fi

if echo "$OUT" | grep -qi "regular directory"; then
    ok "错误信息包含 regular directory"
else
    bad "错误信息缺少 regular directory，实际: $OUT"
fi

if echo "$OUT" | grep -qi "move or rename"; then
    ok "错误信息提示 move or rename"
else
    bad "错误信息缺少 move or rename 引导，实际: $OUT"
fi

# ── 场景 3: 移走目录后 recover 正常工作 ───────────────────────────
log "场景 3: 移走目录后 recover 正常工作"

mv "$WS1" "${WS1}.bak"

set +e
OUT=$("$WS_CKPT" recover -w "$WS1" --force 2>&1)
RC=$?
set -e

if [ "$RC" -eq 0 ]; then
    ok "recover 成功 (rc=0)"
else
    bad "recover 应该成功但 rc=$RC, 输出: $OUT"
fi

if [ -d "$WS1" ]; then
    ok "recover 后路径存在（普通目录）"
else
    bad "recover 后路径不存在"
fi

# recover 后应该变成普通目录而非 symlink
if [ -d "$WS1" ] && [ ! -L "$WS1" ]; then
    ok "recover 后是普通目录（非 symlink）"
else
    bad "recover 后路径类型异常（symlink=$([ -L "$WS1" ] && echo yes || echo no)）"
fi

# ── 场景 4: symlink 不存在时 recover 正常 ─────────────────────────
log "场景 4: symlink 不存在时 recover 正常（路径已删除）"

# 全新 init WS2
mkdir -p "$WS2" && echo "seed" > "$WS2/f.txt"
"$WS_CKPT" init -w "$WS2" >/dev/null 2>&1

# 只删 symlink，不重建
if [ -L "$WS2" ]; then
    rm "$WS2"
else
    bad "WS2 init 后不是 symlink，跳过场景 4"
fi

set +e
OUT=$("$WS_CKPT" recover -w "$WS2" --force 2>&1)
RC=$?
set -e

if [ "$RC" -eq 0 ]; then
    ok "recover 成功 (rc=0)"
else
    bad "recover 应该成功但 rc=$RC, 输出: $OUT"
fi

if [ -d "$WS2" ]; then
    ok "recover 后路径已重建"
else
    bad "recover 后路径不存在"
fi

# ── 汇总 ──────────────────────────────────────────────────────────
echo ""
TOTAL=$((PASS + FAIL))
echo -e "结果: ${PASS}/${TOTAL} passed, ${FAIL} failed"
if [ "$FAIL" -gt 0 ]; then
    echo -e "${RED}VERDICT: FAIL${NC}"
    exit 1
else
    echo -e "${GREEN}VERDICT: PASS${NC}"
    exit 0
fi

Additional Notes