Skills
skills/alemusica/golden-helix-studio/knowledge-research/Gen Agent Trust Hub

knowledge-research

Fail

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill contains hardcoded credentials in the SKILL.md file for accessing a SurrealDB instance. Specifically, it provides Auth: Basic cm9vdDpyb290 which decodes to the default username and password root:root.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes untrusted data from external research sources.
  • Ingestion points: Data is ingested via SQL queries in SKILL.md from the knowledge, chat_message, and paper tables, which contain content from external sources like StackOverflow and GitHub.
  • Boundary markers: No boundary markers or delimiters are defined to separate untrusted database content from system instructions.
  • Capability inventory: The agent uses the retrieved data to research algorithms, debug errors, and find solutions from past conversations, which could lead to the execution of malicious instructions found in those sources.
  • Sanitization: No sanitization, validation, or escaping mechanisms are implemented for the data retrieved from the database before it is processed by the agent.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 1, 2026, 01:03 AM