The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill ingests data from external monitor reports which may contain user-generated or third-party content (e.g., troubleshooting summaries). This data is then used to inform configuration changes.
Ingestion points: The get_monitor_report tool output is consumed in Phase 1 and analyzed in Phase 2 of SKILL.md.
Boundary markers: Absent. The instructions do not include specific delimiters or warnings to ignore potential instructions embedded within the monitor reports.
Capability inventory: The skill has the authority to modify monitor configurations via the create_metric_monitor and create_custom_sql_monitor tools.
Sanitization: Absent. There is no evidence of filtering or escaping logic applied to the content retrieved from the monitor reports before it is used for analysis.
[COMMAND_EXECUTION]: Local File System Interaction. The skill directs the agent to write sensitive monitor reports to a temporary directory on the local file system.
Evidence: Phase 1 of SKILL.md instructs the agent to save reports to /tmp/monitor-report-{monitor_uuid}.md.
Mitigation: The risk of path traversal or injection is mitigated by a strict UUID validation step in Phase 0, ensuring the filename is constrained to a standard format.

tune-monitor