pbvex-svelte
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard CLI tools to inspect the codebase and verify the environment.
- Commands used include
sedfor reading file contents,rg(ripgrep) for searching exported symbols, andpnpmfor running the test suite. - Evidence:
sed -n '1,240p' packages/svelte/package.json,rg -n "export |useQuery|...", andpnpm --filter @pbvex/svelte test. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and interpret source code and configuration files from the local repository, creating a surface for indirect prompt injection.
- Ingestion points: File reads from
packages/svelte/src/index.tsandpackages/svelte/package.jsonviasedandrg. - Boundary markers: There are no explicit instructions or delimiters provided to prevent the agent from following instructions that might be embedded within the source files it reads.
- Capability inventory: The skill has the capability to read files and execute shell commands (
pnpm test). - Sanitization: No sanitization or validation logic is defined for the content extracted from the files before the agent processes them.
Audit Metadata