ably-codebase-review
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill identifies and installs
typescript-language-serverandtypescriptusingnpm install -gif these prerequisites are not found on the system. - [EXTERNAL_DOWNLOADS]: The skill retrieves official error code definitions from Ably's documentation at
https://ably.com/docs/platform/errors/codesto assist in verifying error handling logic. - [INDIRECT_PROMPT_INJECTION]: The skill processes codebase content from various directories, creating a surface for indirect prompt injection where data within the reviewed files could attempt to influence the agent's logic. Ingestion points: Codebase files located in
src/commands/,src/utils/, andtest/unit/commands/. Boundary markers: There are no explicit markers or instructions used to isolate the codebase data from the agent's command logic. Capability inventory: Access to shell commands (npm,grep,which,npx), file system tools (read), theLSPtool, and network requests (WebFetch). Sanitization: The skill does not implement sanitization or filtering of the codebase data before analysis.
Audit Metadata