How do I install the 17 Cowork Bash Mount skill?

Run npx skillmds@latest add kazukinagata/17-cowork-bash-mount in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is 17 Cowork Bash Mount safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with 17 Cowork Bash Mount?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is 17 Cowork Bash Mount free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published 17 Cowork Bash Mount?

kazukinagata (@kazukinagata) published this skill. Their other Agent Skills are listed on their SkillMD profile.

17 Cowork Bash Mount by kazukinagata

Name: 17 Cowork Bash Mount
Author: kazukinagata

17 Cowork Bash Mount

Cowork mounts the plugin install dir under /sessions/CODENAME/mnt/.remote-plugins/ read-only. Bundled scripts can be launched via relative path; CLAUDE_SKILL_DIR expands to a Windows path that fails directly. CLI baseline lets both patterns work. See research section 2.10.

kazukinagata Updated Sep 17, 2026

File contents

17-cowork-bash-mount

§2.10 — bundled script の起動パターン 2 種 + 罠（mkdir -p ゴミ、tee silent fail）。

step 1: alive-check と Pattern A (relative path / find)

proj="${CLAUDE_PROJECT_DIR:-$PWD}"
ver="${VERIFIER_VERSION_DIR:-v-unknown}"
sid="${CLAUDE_SESSION_ID:-no-sid}"
out_dir="$proj/findings/$ver/$sid"
mkdir -p "$out_dir"
{
  printf '[17-BODY %s] tag=alive-check\n' "$(date -Iseconds)"
  printf '[17-BODY] CLAUDE_SKILL_DIR=[%s]\n' "${CLAUDE_SKILL_DIR-(unset)}"
  printf '[17-BODY] CLAUDE_PLUGIN_ROOT=[%s]\n' "${CLAUDE_PLUGIN_ROOT-(unset)}"
} | tee -a "$out_dir/probe.log"

# Pattern A: find then cd
skill_dir=$(find /sessions -path '*/skills/17-cowork-bash-mount' -type d 2>/dev/null | head -1)
if [ -z "$skill_dir" ]; then
  # CLI: SKILL_DIR is a Linux path so use it directly
  skill_dir="${CLAUDE_SKILL_DIR}"
fi
echo "Pattern A: cd $skill_dir && bash scripts/say-hi.sh" | tee -a "$out_dir/probe.log"
( cd "$skill_dir" 2>/dev/null && bash scripts/say-hi.sh ) 2>&1 | tee -a "$out_dir/probe.log"

step 2: Pattern B (${CLAUDE_SKILL_DIR} direct, expected to FAIL on Cowork)

proj="${CLAUDE_PROJECT_DIR:-$PWD}"
ver="${VERIFIER_VERSION_DIR:-v-unknown}"
sid="${CLAUDE_SESSION_ID:-no-sid}"
out_dir="$proj/findings/$ver/$sid"
echo "Pattern B: bash \"${CLAUDE_SKILL_DIR}/scripts/say-hi.sh\"" | tee -a "$out_dir/probe.log"
bash "${CLAUDE_SKILL_DIR}/scripts/say-hi.sh" 2>&1 | tee -a "$out_dir/probe.log" || echo "[17-PATTERN-B exited non-zero (expected on Cowork)]" | tee -a "$out_dir/probe.log"

step 3: 罠の確認 (mkdir -p $DATA gomi + ${CLAUDE_PROJECT_DIR:-$PWD} fallback)

proj="${CLAUDE_PROJECT_DIR:-$PWD}"
ver="${VERIFIER_VERSION_DIR:-v-unknown}"
sid="${CLAUDE_SESSION_ID:-no-sid}"
out_dir="$proj/findings/$ver/$sid"
before_pwd_listing=$(ls "$PWD" 2>/dev/null | head -20)
mkdir -p "${CLAUDE_PLUGIN_DATA}" 2>&1
after_pwd_listing=$(ls "$PWD" 2>/dev/null | head -20)
{
  printf '[17-BODY trap-mkdir] before=%s\n' "$(echo $before_pwd_listing | tr '\n' ' ' | head -c 200)"
  printf '[17-BODY trap-mkdir] after=%s\n' "$(echo $after_pwd_listing | tr '\n' ' ' | head -c 200)"
} | tee -a "$out_dir/probe.log"

完了して exit、./scripts/assert.sh 17。

---name: 17-cowork-bash-mountdescription: Cowork mounts the plugin install dir under /sessions/CODENAME/mnt/.remote-plugins/ read-only. Bundled scripts can be launched via relative path; CLAUDE_SKILL_DIR expands to a Windows path that fails directly. CLI baseline lets both patterns work. See research section 2.10.---# 17-cowork-bash-mount§2.10 — bundled script の起動パターン 2 種 + 罠（`mkdir -p` ゴミ、`tee` silent fail）。## step 1: alive-check と Pattern A (relative path / find)```bashproj="${CLAUDE_PROJECT_DIR:-$PWD}"ver="${VERIFIER_VERSION_DIR:-v-unknown}"sid="${CLAUDE_SESSION_ID:-no-sid}"out_dir="$proj/findings/$ver/$sid"mkdir -p "$out_dir"{ printf '[17-BODY %s] tag=alive-check\n' "$(date -Iseconds)" printf '[17-BODY] CLAUDE_SKILL_DIR=[%s]\n' "${CLAUDE_SKILL_DIR-(unset)}" printf '[17-BODY] CLAUDE_PLUGIN_ROOT=[%s]\n' "${CLAUDE_PLUGIN_ROOT-(unset)}"} | tee -a "$out_dir/probe.log"# Pattern A: find then cdskill_dir=$(find /sessions -path '*/skills/17-cowork-bash-mount' -type d 2>/dev/null | head -1)if [ -z "$skill_dir" ]; then # CLI: SKILL_DIR is a Linux path so use it directly skill_dir="${CLAUDE_SKILL_DIR}"fiecho "Pattern A: cd $skill_dir && bash scripts/say-hi.sh" | tee -a "$out_dir/probe.log"( cd "$skill_dir" 2>/dev/null && bash scripts/say-hi.sh ) 2>&1 | tee -a "$out_dir/probe.log"```## step 2: Pattern B (${CLAUDE_SKILL_DIR} direct, expected to FAIL on Cowork)```bashproj="${CLAUDE_PROJECT_DIR:-$PWD}"ver="${VERIFIER_VERSION_DIR:-v-unknown}"sid="${CLAUDE_SESSION_ID:-no-sid}"out_dir="$proj/findings/$ver/$sid"echo "Pattern B: bash \"${CLAUDE_SKILL_DIR}/scripts/say-hi.sh\"" | tee -a "$out_dir/probe.log"bash "${CLAUDE_SKILL_DIR}/scripts/say-hi.sh" 2>&1 | tee -a "$out_dir/probe.log" || echo "[17-PATTERN-B exited non-zero (expected on Cowork)]" | tee -a "$out_dir/probe.log"```## step 3: 罠の確認 (mkdir -p $DATA gomi + ${CLAUDE_PROJECT_DIR:-$PWD} fallback)```bashproj="${CLAUDE_PROJECT_DIR:-$PWD}"ver="${VERIFIER_VERSION_DIR:-v-unknown}"sid="${CLAUDE_SESSION_ID:-no-sid}"out_dir="$proj/findings/$ver/$sid"before_pwd_listing=$(ls "$PWD" 2>/dev/null | head -20)mkdir -p "${CLAUDE_PLUGIN_DATA}" 2>&1after_pwd_listing=$(ls "$PWD" 2>/dev/null | head -20){ printf '[17-BODY trap-mkdir] before=%s\n' "$(echo $before_pwd_listing | tr '\n' ' ' | head -c 200)" printf '[17-BODY trap-mkdir] after=%s\n' "$(echo $after_pwd_listing | tr '\n' ' ' | head -c 200)"} | tee -a "$out_dir/probe.log"```完了して exit、`./scripts/assert.sh 17`。

Frequently asked questions