linear-data-handling

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPERSISTENCE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes user-generated content (issue titles and descriptions) from the Linear API and webhooks. This content is stored in a local database and included in backups without specific sanitization for natural language instructions, potentially influencing an agent that later processes this data.
  • Ingestion points: Data is retrieved via client.issues and webhook payloads in both SKILL.md and references/implementation.md.
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore potentially malicious content within these fields.
  • Capability inventory: The skill utilizes fs.writeFile to create backup files and extensive database write operations.
  • Sanitization: While zod is used for structural validation, there is no logic to filter or escape instructions embedded within string fields like description.
  • [PERSISTENCE]: The skill implements a persistence mechanism via cron.schedule in references/implementation.md. This is used to maintain long-term execution for periodic data consistency checks and synchronization tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 08:05 PM
Security Audit — agent-trust-hub — linear-data-handling