threads-growth-skill
Fail
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: HIGHPERSISTENCECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONTIME_DELAYED_CONDITIONAL
Full Analysis
- [REMOTE_CODE_EXECUTION]: The installation process involves cloning a repository from an unverified source (github.com/krumjahn/threads-growth-skill) and executing a shell script (scripts/setup.sh) from that repository.\n- [PERSISTENCE]: The skill configures a bi-weekly cron job (0 2 * * 0 ...) to automatically run the scrape_insights.sh script in the background, establishing a permanent task on the user's system.\n- [CREDENTIALS_UNSAFE]: Documentation instructs the user to export active session cookies from their browser and save them in plain text to cookies.json. These cookies provide full authenticated access to the user's Threads account.\n- [EXTERNAL_DOWNLOADS]: The skill fetches code and references from GitHub accounts (krumjahn, akseolabs-seo) that are not part of the official resource list for the vendor.\n- [COMMAND_EXECUTION]: The skill relies on local shell scripts (setup.sh, scrape_insights.sh) for setup, authentication verification, and automated data collection.\n- [TIME_DELAYED_CONDITIONAL]: The automated background scraping task is governed by a time-based condition to execute at specific intervals.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests raw post text from the Threads platform, which is processed to identify patterns and generate drafts.\n
- Ingestion points: scrape_insights.sh retrieves text content from the user's recent posts.\n
- Boundary markers: No specific delimiters or "ignore" instructions are provided to isolate the scraped content from the system prompt.\n
- Capability inventory: The skill possesses capabilities for shell execution, file system modification, and cron job management.\n
- Sanitization: No sanitization or safety checks are described for the external content before it is used in prompt generation.
Recommendations
- AI detected serious security threats
Audit Metadata