twitter-GTM-find-Skill

Pass

Audited by Gen Agent Trust Hub on Jun 11, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted content (tweet text and author bios) directly into LLM prompts for extraction and verification.
  • Ingestion points: Data enters the pipeline via the Twitter scraper in scripts/src/index.ts.
  • Boundary markers: The prompts in scripts/src/extractor.ts and scripts/src/icp-filter.ts use basic double-quote encapsulation for interpolated data, which provides minimal protection against adversarial input.
  • Capability inventory: The skill possesses the ability to write to the local file system (radar-jobs.json, openclaw-icp-jobs.json) and access live web data via Gemini's Google Search grounding tool.
  • Sanitization: No explicit sanitization or filtering of the scraped tweet content is performed before it is passed to the LLM.
  • [EXTERNAL_DOWNLOADS]: The skill relies on an external service (Apify) to execute a remote scraping actor (ghSpYIW3L1RvT57NT). This involves outbound network requests to a third-party platform to perform its primary data collection task.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 11, 2026, 05:29 PM
Security Audit — agent-trust-hub — twitter-GTM-find-Skill