crypto-daily

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes bash commands and shell scripts to manage local files, determine dates, and extract specific content from markdown reports for processing.
  • [COMMAND_EXECUTION]: Interacts with external CLI tools (telegram-cli.py and chrome-use) via the shell to post updates to Telegram channels and X.com profiles.
  • [DYNAMIC_EXECUTION]: Dynamically generates and executes a Python script to parse JSON output from an AI subagent and append recorded scores to a local CSV file (.cache/crypto-advisor/crypto-advisor.eval.csv).
  • [PROMPT_INJECTION]: Presents an indirect prompt injection surface by reading and processing data from research/crypto-portfolio-${TODAY}.md. The skill mitigates this by using specific delimiters and extracting only predefined sections for publication and evaluation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 02:21 PM
Security Audit — agent-trust-hub — crypto-daily