Cancorteaw App

ReviewAudited by ClawScan on May 10, 2026.

Overview

The app-building workflow is plausible, but the skill’s main action runs an unprovided local shell script, so its safety claims cannot be verified from the submitted artifacts.

Review /home/patron/apps/_bin/appctl before installing or invoking this skill. Confirm it really restricts work to /home/patron/apps, validates arguments, and only runs the documented Expo commands. Also be aware that it may run npm/Expo tooling and leave a local preview process running until you stop it.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Installing the skill may cause the agent to run whatever script exists at /home/patron/apps/_bin/appctl on the host, which could behave differently from the documentation.

Why it was flagged

The skill’s primary execution target is a local script outside the submitted package. No appctl source or install spec is provided, so the reviewer cannot verify the claimed allowlist or path restrictions.

Skill content
"entrypoint": { "type": "shell", "command": "/home/patron/apps/_bin/appctl", "args": [ "{{args}}" ] }
Recommendation

Include the appctl source in the skill package or provide a pinned, auditable install method/checksum; users should inspect that script before enabling the skill.

What this means

If the local runner is not strict, agent-supplied arguments could lead to commands or file changes beyond the intended Expo workflow.

Why it was flagged

The skill frames the runner as controlled, but the included artifacts do not show the enforcement logic. The allowed tools also include broad interpreters/package executors such as bash, python3, and npx.

Skill content
The runner is **allowlisted**: only `node`, `npm`, `npx`, `git`, `bash`, `python3` can be invoked.
Recommendation

Verify that appctl validates subcommands, project names, paths, environment variables, and child commands before use; prefer explicit allowlisted argument parsing in reviewed code.

What this means

Creating or previewing apps may run npm/Expo tooling on the local machine.

Why it was flagged

npx-based Expo scaffolding can download and execute npm tooling. This is expected for an Expo app-builder skill and is disclosed, but it is still code execution from the npm ecosystem.

Skill content
Uses `npx create-expo-app` in non-interactive mode
Recommendation

Use only in the intended workspace, consider pinning package versions, and avoid running it on sensitive projects without reviewing the npm tooling.

What this means

A local Expo preview server may keep running, using a port and system resources after the command finishes.

Why it was flagged

The preview command creates a long-running local process and stores its pid. This persistence is disclosed and purpose-aligned, but it remains active until stopped.

Skill content
Writes pid to: `/home/patron/apps/_state/<name>.pid` ... To stop preview: `kill $(cat /home/patron/apps/_state/<name>.pid)`
Recommendation

Stop previews when done and consider adding a documented `stop` command rather than relying on manual kill commands.