The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.Popen in db-load-xml.py and Invoke-PlatformProcess in db-load-xml.ps1 to execute the 1C:Enterprise platform executables (1cv8.exe or ibcmd.exe). This is the primary and intended function of the skill to perform database configuration loads. The scripts include robust sanitization for paths and arguments to prevent command injection.
[INDIRECT_PROMPT_INJECTION]: The skill processes external XML configuration files and .v8-project.json project metadata.
Ingestion points: Reads XML source files from a directory specified by the user and parses .v8-project.json for database connection details.
Boundary markers: The skill output is clearly labeled with "--- Log ---" and "--- Вывод платформы ---" to help the agent distinguish between tool output and instructions.
Capability inventory: The skill can execute local platform binaries and write temporary log files in the system temp directory.
Sanitization: The skill validates file paths and employs strict argument parsing. It does not interpolate the content of the XML files directly into prompts, but rather reports the status of the loading operation.
[CREDENTIALS_SAFE]: The skill handles 1C database credentials. It includes specific logic (Protect-Secrets in PowerShell and _redact in Python) to ensure passwords and sensitive tokens are replaced with asterisks before being printed to the console or logs.

db-load-xml