clawl skill
Analysis
This skill mostly matches its registration purpose, but it sends agent metadata to an unexpected default host and overwrites a local discovery file without the confirmation its documentation promises.
Findings (3)
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.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
const CLAWL_API = process.env.CLAWL_API || 'https://moogle-alpha.vercel.app';
The documentation says the script pings clawl.co.uk, but the bundled script defaults to a different Vercel host for API calls. That is a material mismatch in where registration metadata is sent.
const outputPath = path.join(process.cwd(), 'clawl.json'); fs.writeFileSync(outputPath, JSON.stringify(clawlJson, null, 2));
The implementation writes clawl.json directly in the current workspace. This conflicts with the SKILL.md error-handling claim that an existing clawl.json will be shown and confirmed before overwriting.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
const soulPaths = [
path.join(process.cwd(), 'SOUL.md'),
path.join(process.env.HOME || process.env.USERPROFILE || '', 'clawd', 'SOUL.md'),
];The script reads persistent local identity/context files such as SOUL.md and IDENTITY.md to derive public registration metadata. This is aligned with the skill's purpose, but those files may contain identity details the user did not intend to publish.
