The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The installation instructions explicitly require the use of sudo to run make install, which grants the skill installation process full administrative access to the system.
[COMMAND_EXECUTION]: The documentation indicates that the logfile argument can accept a shell command as input. If an agent passes unvalidated user input into this argument, it allows for arbitrary command execution on the host system.
[METADATA_POISONING]: The skill's YAML frontmatter contains the field verification: "security_reviewed". This is a self-asserted claim of safety designed to mislead automated scanners or users into trusting the content without independent verification.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external PostgreSQL logs, which are untrusted data sources.
Ingestion points: Reads raw log files or shell command outputs containing database logs in SKILL.md.
Boundary markers: There are no instructions provided to the agent to treat the parsed log content as untrusted or to use delimiters to prevent log data from being interpreted as instructions.
Capability inventory: The skill has the ability to execute shell commands and read local files.
Sanitization: No sanitization or validation of the log content is described before the agent processes it.

Turn PostgreSQL logs into incident and tuning reports with pgBadger