alertmanager-query
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection via API responses.
- Ingestion points: JSON responses from AlertManager and VictoriaMetrics API endpoints (SKILL.md, references/api-reference.md).
- Boundary markers: Absent.
- Capability inventory: Shell command execution via curl, jq, and date (SKILL.md).
- Sanitization: Present; the skill uses jq --arg for safe interpolation of user-supplied data into requests (SKILL.md).
- [COMMAND_EXECUTION]: The skill utilizes curl, jq, and date to interact with APIs and process monitoring data. Evidence: Frontmatter allowed-tools and documented usage examples in SKILL.md.
- [CREDENTIALS_UNSAFE]: The skill references a local configuration file for authentication (~/.config/victoriametrics/curl.conf). This is documented as a secure best practice to avoid exposing sensitive headers in command history or process listings. Evidence: SKILL.md 'Auth Pattern' section.
Audit Metadata