algolia-incident-runbook
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data including user-reported symptoms, error logs, and external search query results during the incident response process.
- Ingestion points: Symptom descriptions, provider status messages, and query responses from external Algolia indices (SKILL.md).
- Boundary markers: Instructions explicitly require comparing application results against a known "source-of-truth" to validate data integrity.
- Capability inventory: The skill is granted access to file system read/write tools and web fetching capabilities for documentation.
- Sanitization: The runbook mandates verification using synthetic queries and representative user data to confirm recovery and relevance before closing an incident.
- [EXTERNAL_DOWNLOADS]: The skill references and fetches information from official Algolia domains.
- Evidence: References to status.algolia.com and algolia.com for official documentation and API monitoring.
- Verification: The target domains are the official resources of the service provider identified in the skill's primary purpose.
Audit Metadata