wrangler
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill contains hardcoded Cloudflare account identifiers and an email address (Steipete@gmail.com). These correspond to the author's own environment ('steipete') and are documented as vendor-owned resources.- [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill instructions use 'npm exec' to download and run the 'wrangler' CLI package from the NPM registry at runtime. As Cloudflare is a well-known service, this dependency is considered safe.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, which presents an injection surface.
- Ingestion points: Data is ingested through 'wrangler tail' (live application logs), 'wrangler kv key get' (remote database values), and reading local 'wrangler.toml' or 'wrangler.json' configuration files.
- Boundary markers: The instructions lack explicit delimiters or warnings to the agent to ignore instructions that might be embedded in the ingested data.
- Capability inventory: The skill grants the agent the ability to execute shell commands, deploy code to live environments, and modify remote data stores.
- Sanitization: There are no instructions for validating or sanitizing the output of the CLI tools before the agent interprets or acts upon the content.
Audit Metadata