java-architect
Fail
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: CRITICALEXTERNAL_DOWNLOADSMETADATA_POISONINGCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill directs users and the agent to documentation at
https://jeffallan.github.io/claude-skills/skills/language/java-architect/, which has been flagged as malicious by multiple reputation scanners. - [METADATA_POISONING]: The skill metadata (SKILL.md) itself has been flagged by file reputation scanners for malicious patterns. The inclusion of external links to a domain with a confirmed malicious reputation suggests a high risk to the execution environment.
- [COMMAND_EXECUTION]: The workflow mandates the execution of build tool commands such as
./mvnw verifyand./gradlew check. While standard in Java development, these scripts execute local code and should be carefully audited given the suspicious reputation of the skill's source metadata. - [INDIRECT_PROMPT_INJECTION]: The skill establishes a vulnerability surface by processing untrusted data (external source code, SQL logs, and JaCoCo HTML reports) and using that data to inform further actions.
- Ingestion points: Project source code, Hibernate SQL logs, and JaCoCo reports located at
target/site/jacoco/index.html. - Boundary markers: No delimiters or protective instructions are used to separate untrusted external content from agent instructions.
- Capability inventory: The agent is authorized to execute file system operations and build commands based on the ingested data.
- Sanitization: No validation or filtering mechanisms are present for the ingested HTML or log files.
Recommendations
- CRITICAL: 1 file(s) identified as malware by FileRep - DO NOT USE
- AI detected serious security threats
- Contains 2 malicious URL(s) - DO NOT USE
Audit Metadata