twitter-automation
Pass
Audited by Gen Agent Trust Hub on Nov 18, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted content (such as text for tweets or direct messages) and uses it to perform actions via the Twitter API.
- Ingestion points: Input data is accepted via JSON objects in the
belt app runcommands (e.g., in SKILL.md). - Boundary markers: The skill uses standard JSON formatting for tool inputs, which helps delimit data from instructions.
- Capability inventory: The skill allows posting, liking, retweeting, sending DMs, and following users via the
beltCLI tool which performs network operations. - Sanitization: There are no explicit instructions for the agent to sanitize or validate the content of the tweets or messages before execution.
- [EXTERNAL_DOWNLOADS]: The skill documentation encourages the installation of external tools and additional skill packages from the inference-sh organization.
- Evidence: The skill references a CLI installation script hosted at
https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.mdand usesnpx skills addto fetch remote skill definitions.
Audit Metadata