tencent-news

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the tencent-news-cli executable and corresponding checksum files from Tencent's verified content delivery network (mat1.gtimg.com) during initial setup and version updates.
  • [REMOTE_CODE_EXECUTION]: Instructions and referenced guides provide commands to execute remote installation and setup scripts from well-known domains belonging to Tencent and the Bun runtime.
  • [COMMAND_EXECUTION]: The skill utilizes shell and TypeScript scripts to manage binary permissions and invoke CLI subcommands for system status, API key retrieval, and news search.
  • [PROMPT_INJECTION]: An indirect prompt injection surface is present as the skill processes external news data. 1. Ingestion points: News titles, sources, and summaries retrieved from the Tencent News CLI. 2. Boundary markers: Absent. 3. Capability inventory: The agent can execute shell scripts and the local CLI binary. 4. Sanitization: Content is parsed via JSON and formatted according to Markdown templates, though no specific filtering for adversarial instructions is performed on the raw news strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 07:33 AM
Security Audit — agent-trust-hub — tencent-news