newsletter-curation

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions utilize the belt CLI tool via the Bash tool to perform searches, process images, and interact with social media APIs. This behavior is consistent with the skill's stated purpose of automating newsletter curation tasks.
  • [EXTERNAL_DOWNLOADS]: The skill references and provides instructions for installing external dependencies, including the belt-sh/cli and various sub-skills from the inference-sh repository (e.g., inference-sh/skills@email-design). These are presented as necessary components for the newsletter curation workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for ingesting untrusted data from external sources via search tools (tavily/search-assistant and exa/search). This content is subsequently processed into newsletter summaries and commentary, creating a potential surface for indirect prompt injection if the source data contains malicious instructions.
  • Ingestion points: External search results from Tavily and Exa are ingested in SKILL.md.
  • Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the search results.
  • Capability inventory: The skill possesses the capability to execute shell commands via Bash(belt *) and perform network operations.
  • Sanitization: No sanitization or filtering logic is documented for the ingested search content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 01:13 PM
Security Audit — agent-trust-hub — newsletter-curation