sillytavern-api-reference
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides navigation links to several GitHub repositories (SillyTavern/SillyTavern, N0VI028/JS-Slash-Runner, zonde306/ST-Prompt-Template, MagicalAstrogy/MagVarUpdate) to serve as authoritative sources for API verification. These are the standard community repositories for the tools described.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface where it processes data related to SillyTavern rolecards, which often ingest untrusted user chat data. However, the skill explicitly provides defensive instructions, such as:
- Warning users/agents to "Never expose API keys, authentication data, extension settings, or private chat content."
- Requiring explicit verification of "remote runtime code" delivered via Git or CDN.
- Using capability probes and narrow API implementations to prevent accidental execution of unauthorized commands.
- [SAFE]: The skill's primary purpose is defensive documentation and version verification. It does not perform any autonomous network operations, command executions, or data exfiltration. All identified external links are used for documentation reference rather than automated execution.
Audit Metadata