Back to skill

Security audit

desktop-pilot

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because the scanner found environment-related concerns and could not fully inspect several referenced components, leaving the evidence incomplete.

Treat the suspicious verdict as a reason to review carefully before installing. Several files were only partially inspected, and the flagged permissions are not known to be enforced by the host, so these results do not establish that the skill is safe.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Rogue AgentSelf-Modification, Session Persistence
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (19)

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 50)May include surrounding context.

md
`scripts/browser/overlay.js` draws a glow and status pill only. Inject it with your existing browser automation tool, then use `window.__dpOverlay.show('text')`

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · SKILL.md (reported line 52)May include surrounding context.

md
### Informational overlay inside a browser page (any OS)
`scripts/browser/overlay.js` draws a glow and status pill only. Inject it with your existing browser automation tool, then use `window.__dpOverlay.show('text')`, `.text('...')`, `.hide()` and `.visible()`.

**Breaking change in 1.0.2:** the page overlay no longer has a Stop button or `stopped()`/`clear()` methods. An untrusted page can change its own scripts and DOM, so it must not own cancellation or permission to resume. Cancel using the trusted agent host's control outside the webpage; on Windows/WSL the native desktop overlay provides Stop. Maintain host-side cancellation across navigation and before every tool call. The browser overlay is not an authorization signal, even though its API object is frozen. If your host has no trusted cancellation control, do not run unattended browser tasks.

## Platforms
- **CLI requirement:** Bash 4+ (WSL/Linux or Git Bash/MSYS on Windows). macOS ships Bash 3.2, so use a modern Bash for `des

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

md
x or Git Bash/MSYS on Windows). macOS ships Bash 3.2, so use a modern Bash for `desktop.sh`, or invoke the built-in JXA backend directly. No automatic dependenc

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 74)May include surrounding context.

md
sistent Stop file under the current Windows user's temporary directory. Native `desktop.ps1`, the Bash wrapper, and overlay launch all check it. Typing checks i

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 76)May include surrounding context.

md
r-overlay.test.js` for the informational browser UI. Live Windows tests are in `tests/windows-regression.ps1`; they operate a dedicated scratch window and isola

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 83)May include surrounding context.

sh
overlay_hook "$OV_L"  # start is not authorization to resume after Stop
      cp "$HERE/windows/overlay.ps1" "$OV_L/../overlay.ps1"
      cp "$HERE/windows/launch-overlay.ps1" "$OV_L/../launch-overlay.ps1"
      rm -f "$OV_L/quit.flag"
      if [[ -n "${O[lock]:-}" && "${O[lock]}" != 0 ]]; then touch "$OV_L/lock.flag"; else rm -f "$OV_L/lock.flag"; fi
      printf '%s' "${O[text]:-${DESKTOP_PILOT_OVERLAY_TEXT:-Agent is working...}}" > "$OV_L/text.txt"
      touch "$OV_L/beat"

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 84)May include surrounding context.

sh
cp "$HERE/windows/overlay.ps1" "$OV_L/../overlay.ps1"
      cp "$HERE/windows/launch-overlay.ps1" "$OV_L/../launch-overlay.ps1"
      rm -f "$OV_L/quit.flag"
      if [[ -n "${O[lock]:-}" && "${O[lock]}" != 0 ]]; then touch "$OV_L/lock.flag"; else rm -f "$OV_L/lock.flag"; fi
      printf '%s' "${O[text]:-${DESKTOP_PILOT_OVERLAY_TEXT:-Agent is working...}}" > "$OV_L/text.txt"
      touch "$OV_L/beat"
      if overlay_alive; then echo "OK overlay already running (text updated)"; return; fi

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 88)May include surrounding context.

sh
printf '%s' "${O[text]:-${DESKTOP_PILOT_OVERLAY_TEXT:-Agent is working...}}" > "$OV_L/text.txt"
      touch "$OV_L/beat"
      if overlay_alive; then echo "OK overlay already running (text updated)"; return; fi
      rm -f "$OV_L/pid"
      local launcher="${OV_WTEMP}desktop-pilot\\launch-overlay.ps1"
      local launchargs=(-Dir "$OV_W" -IdleSeconds "${O[idle]:-120}" -LockIdleSeconds "${O[lock-idle]:-8}")
      [[ "${O[capture]:-0}" == 1 ]] && launchargs+=(-VisibleInCapture)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 103)May include surrounding context.

sh
if overlay_alive; then touch "$OV_L/quit.flag"; local i; for i in $(seq 1 15); do overlay_alive || break; sleep 0.3; done
        if overlay_alive; then die "Overlay has not exited yet; quit request retained"; fi
        echo "OK overlay stopped"
      else rm -f "$OV_L/pid"; echo "OK overlay not running"; fi
      rm -f "$OV_L/quit.flag" ;;
    clear) rm -f "$OV_L/stop.flag"; echo "OK stop flag cleared" ;;
    lock)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 104)May include surrounding context.

sh
if overlay_alive; then die "Overlay has not exited yet; quit request retained"; fi
        echo "OK overlay stopped"
      else rm -f "$OV_L/pid"; echo "OK overlay not running"; fi
      rm -f "$OV_L/quit.flag" ;;
    clear) rm -f "$OV_L/stop.flag"; echo "OK stop flag cleared" ;;
    lock)
      overlay_hook "$OV_L"

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 105)May include surrounding context.

sh
echo "OK overlay stopped"
      else rm -f "$OV_L/pid"; echo "OK overlay not running"; fi
      rm -f "$OV_L/quit.flag" ;;
    clear) rm -f "$OV_L/stop.flag"; echo "OK stop flag cleared" ;;
    lock)
      overlay_hook "$OV_L"
      overlay_alive || die "overlay is not running (overlay start first)"

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 110)May include surrounding context.

sh
overlay_hook "$OV_L"
      overlay_alive || die "overlay is not running (overlay start first)"
      touch "$OV_L/lock.flag" "$OV_L/beat"; echo "OK input lock requested (engages with your next action, releases 8 s after the last one; Esc held 1.5 s or Stop releases it)" ;;
    unlock) rm -f "$OV_L/lock.flag"; echo "OK input lock released" ;;
    auto)
      local cfgdir="${XDG_CONFIG_HOME:-$HOME/.config}/desktop-pilot"
      case "$ARG" in

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
95% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · scripts/desktop.sh (reported line 116)May include surrounding context.

sh
case "$ARG" in
        on) mkdir -p "$cfgdir"; printf '%s\n' "${O[text]:-${DESKTOP_PILOT_OVERLAY_TEXT:-Agent is working...}}" > "$cfgdir/auto-overlay"
            echo "OK auto overlay ON: the overlay now starts by itself before any action (text: $(head -1 "$cfgdir/auto-overlay")); it closes on its own when idle" ;;
        off) rm -f "$cfgdir/auto-overlay"; echo "OK auto overlay OFF" ;;
        *) if [[ -f "$cfgdir/auto-overlay" ]]; then echo "auto overlay: on ($(head -1 "$cfgdir/auto-overlay"))"; else echo "auto overlay: off (enable with: overlay auto on [--text S])"; fi ;;
      esac ;;
    selftest)

Code capability 'env' detected in tests/browser-live.test.js but not covered by declared permissions.

High
Category
MCP Least Privilege
Confidence
75% confidence
Finding

The skill uses 'env' capability that is not covered by the declared 'allowed-tools' preapproval. Review the host runtime's actual enforcement policy; runtime permissions are unknown unless separately verified, and this declaration alone does not establish a permission ceiling.

Content

No source excerpt is available for this finding.

Code capability 'file_read' detected in tests/test_regressions.py but not covered by declared permissions.

High
Category
MCP Least Privilege
Confidence
55% confidence
Finding

The skill uses 'file_read' capability that is not covered by the declared 'allowed-tools' preapproval. Review the host runtime's actual enforcement policy; runtime permissions are unknown unless separately verified, and this declaration alone does not establish a permission ceiling.

Content

No source excerpt is available for this finding.

Code capability 'file_write' detected in tests/test_regressions.py but not covered by declared permissions.

High
Category
MCP Least Privilege
Confidence
55% confidence
Finding

The skill uses 'file_write' capability that is not covered by the declared 'allowed-tools' preapproval. Review the host runtime's actual enforcement policy; runtime permissions are unknown unless separately verified, and this declaration alone does not establish a permission ceiling.

Content

No source excerpt is available for this finding.

Code capability 'shell' detected in tests/test_regressions.py but not covered by declared permissions.

High
Category
MCP Least Privilege
Confidence
55% confidence
Finding

The skill uses 'shell' capability that is not covered by the declared 'allowed-tools' preapproval. Review the host runtime's actual enforcement policy; runtime permissions are unknown unless separately verified, and this declaration alone does not establish a permission ceiling.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · tests/test_regressions.py (reported line 42)May include surrounding context.

python
'DESKTOP_PILOT_AUTO_OVERLAY': '0'}

    def run_cli(self, platform, *args):
        return subprocess.run(['bash', str(ROOT/'scripts/desktop.sh'), *args],
                              env=dict(self.env, FAKE_OS=platform), capture_output=True, text=True)

    def test_stop_survives_all_restart_and_action_paths(self):

Scope Creep

Low
Category
Excessive Agency
Confidence
70% confidence
Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · LICENSE (reported line 7)May include surrounding context.

text
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.