build-mcp-server-sdk-v2
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting and analyzing user project files, such as
package.jsonand source code, to identify SDK versions and provide implementation guidance. This creates a potential surface where an attacker could embed malicious instructions within project files to influence the agent's actions during the analysis or development workflow. - Ingestion points:
SKILL.mddirects the agent to inspect the project structure usingtree, and read the contents ofpackage.jsonand files within thesrc/directory to identify version fingerprints. - Capability inventory: The agent is instructed to perform tasks including installing npm packages, running local shell scripts (such as the provided version checker), and modifying project source code.
- Boundary markers: The skill includes specific fingerprinting tables and "Core rules" to guide the agent's analysis, but it does not provide explicit directives to ignore potential instructions embedded within the processed project data.
- Sanitization: No specific sanitization, escaping, or validation of the content read from project files is described before it is interpolated into the agent's context for decision-making and command generation.
Audit Metadata