map-the-landscape
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
README.mdprovides installation instructions usingnpx @tamng0905/builder-essential-skills, which involves downloading and executing a package from the npm registry.- [COMMAND_EXECUTION]: TheSKILL.mdinstructions direct the agent to execute shell commands such asrg --files(ripgrep) to gather information about the repository structure and entry points.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data without adequate safety constraints. - Ingestion points: The agent reads local repository files (source code, manifests, documentation) and browses external web sources to gather evidence for the landscape map.
- Boundary markers: The instructions do not specify delimiters (e.g., XML tags) or include warnings for the agent to ignore any instructions embedded within the analyzed files or web content.
- Capability inventory: The agent has the capability to read files, execute shell commands for file discovery, access the internet for browsing, and write its findings to the local filesystem as a markdown artifact.
- Sanitization: There are no protocols defined for the agent to sanitize or escape the content it ingests before using it for reasoning or including it in the final output.
Audit Metadata