linear-security-basics
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from a user-provided
[repository-path]to identify security gaps. - Ingestion points: Files within the provided repository path are read using
Read,Glob, andGreptools. - Boundary markers: The instructions do not define explicit delimiters or "ignore embedded instructions" wrappers when processing the content of files from the repository.
- Capability inventory: The skill has
WriteandEdittools to modify the repository, andWebFetchto access external documentation. If the agent is influenced by malicious instructions hidden in the code it reviews, it could potentially perform unauthorized edits or leak data. - Sanitization: The skill includes strong negative constraints, such as "Never write credentials, customer content, unrestricted environment output, or unredacted GraphQL variables" and explicit requirements to redact sensitive fields from telemetry.
Audit Metadata