write-coding-standards-from-file
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection attack surface because it is designed to ingest and process the contents of arbitrary user-provided files and folders.
- Ingestion points: The
${fileName}and${folderName}parameters allow the agent to read local file content. - Boundary markers: The skill lacks explicit boundary markers or instructions to treat ingested file content strictly as data, increasing the risk that embedded instructions in source code (e.g., in comments) could be obeyed by the agent.
- Capability inventory: The agent has the ability to create new files (
createNewFile), modify existing files (addToREADME), and perform network operations (fetchStyleURL). - Sanitization: No sanitization or filtering is specified for the data extracted from the analyzed files.
- [EXTERNAL_DOWNLOADS]: The skill includes a configuration to fetch data from a long list of external URLs to provide context for coding standards.
- Evidence: The
### Fetch Linkssection includes over 30 URLs for various language style guides. - Analysis: The links target reputable and well-known documentation sources, including official language repositories (Python PEPs, Rust, Swift), cloud provider documentation (Microsoft), and established community resources (W3Schools, SQLStyle.guide, GitHub). These fetches are intended for context gathering and do not involve direct execution of remote scripts.
Audit Metadata