The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/mine_error_log.py script utilizes subprocess.run to call the curl binary. The URL passed to this command is derived directly from user-supplied command-line arguments, facilitating dynamic network-based command execution.\n- [DATA_EXFILTRATION]: The skill is purpose-built to harvest sensitive data from external systems. It employs regular expressions to search for and extract database connection strings (DB_USER, DB_PASSWORD, DB_HOST), various API keys (Stripe, Google, AWS, JWT), and internal IP addresses. The harvested data is systematically saved into local files (intel_summary.md and JSON reports).\n- [PROMPT_INJECTION]: The skill ingests untrusted content from remote PHP error logs and presents this data to the agent for analysis, which represents a surface for Indirect Prompt Injection.\n
Ingestion points: External error_log files are fetched via curl and read into memory in both the shell procedures within SKILL.md and the Python analysis script scripts/mine_error_log.py.\n
Boundary markers: There are no explicit boundary markers or instructions to the agent to distinguish between data and potential commands within the analyzed logs.\n
Capability inventory: The skill has capabilities including shell command execution (curl, grep), file system modification (mkdir, writing .md and .json files), and network connectivity.\n
Sanitization: The skill does not perform sanitization or safety filtering on the extracted log content before it is processed by the agent.

error-log-mining