neta-character

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text for character definitions (names, descriptions, personas, and triggers) which are subsequently utilized by AI models for content generation and interaction. This creates an attack surface for indirect prompt injection if malicious instructions are embedded in these metadata fields.\n
  • Ingestion points: Arguments such as --description, --persona, and --trigger passed to the CLI tool as described in SKILL.md and references/character-creation.md.\n
  • Boundary markers: None defined within the command templates or instructions.\n
  • Capability inventory: Character management operations via the @talesofai/neta-skills CLI tool.\n
  • Sanitization: No sanitization or validation processes for user-provided text are documented.\n- [COMMAND_EXECUTION]: The skill instructs the agent to perform character management operations by executing the @talesofai/neta-skills CLI tool.\n- [EXTERNAL_DOWNLOADS]: The skill downloads and executes the @talesofai/neta-skills package from the NPM registry, which provides the functional interface for the Neta platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 09:53 AM