awsclaw-s3

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by combining data ingestion with high-privilege capabilities.
  • Ingestion points: The GetObject command (using AsText: true) and SelectObjectContent read data from S3 objects directly into the agent's active context.
  • Boundary markers: There are no instructions or delimiters defined to prevent the agent from following instructions embedded within the S3 object data.
  • Capability inventory: The toolset includes powerful operations such as PutBucketPolicy, DeleteBucket, DeleteObject, and PutObject which could be misused if the agent is influenced by malicious data.
  • Sanitization: No sanitization or validation mechanisms are specified for the data retrieved from external objects.
  • [DATA_EXFILTRATION]: The GetObject and SelectObjectContent commands allow for the extraction of sensitive data from private S3 storage into the agent's context, where it could potentially be processed or moved.
  • [COMMAND_EXECUTION]: The skill exposes a wide array of AWS S3 API operations, granting the agent significant control over cloud storage infrastructure, including the ability to modify access control lists and encryption settings.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 08:11 PM
Security Audit — agent-trust-hub — awsclaw-s3