set-zeabur-conventions
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the official
@zeabur/mcp-serverusingnpxfrom the NPM registry. Zeabur is a well-known PaaS platform, and the package originates from their official scope. - [INDIRECT_PROMPT_INJECTION]: The skill reads and modifies project files (
AGENTS.mdandCLAUDE.md) which could contain untrusted data. - Ingestion points: Reads existing content from
AGENTS.mdin the project root to check for version consistency. - Boundary markers: Uses a specific section header
## Zeabur 部署規範to delimit its modifications. - Capability inventory: Performs file system read and write operations.
- Sanitization: The skill identifies existing sections to avoid duplicate writes but does not perform complex sanitization on pre-existing text.
- [CREDENTIALS_UNSAFE]: While the skill mentions the
ZEABUR_TOKENenvironment variable, it uses standard placeholders (sk-xxxxxx) and includes explicit instructions for the AI to warn users about token exposure, advocating for revocation if secrets are leaked in conversation history.
Audit Metadata