apollo-federation
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
roverCLI tool (specificallyrover supergraph composeandrover subgraph check) to perform schema validation and composition tasks. These commands are restricted to the official Apollo GraphQL CLI toolset, which is consistent with the skill's purpose. - [EXTERNAL_DOWNLOADS]: The skill references several external URLs for schema specifications and documentation, such as
https://specs.apollo.dev/federation/v2.12andhttps://www.apollographql.com/docs/.... These are authoritative resources provided by the skill author (apollographql) and are used correctly within the Federation SDL context. - [INDIRECT_PROMPT_INJECTION]: The skill provides a framework for the agent to ingest and analyze GraphQL Schema Definition Language (SDL) data. While SDL files could theoretically contain untrusted text in descriptions or comments, the skill's primary focus is on formal composition validation via the
roverCLI, which limits the potential for exploitation. - Ingestion points: GraphQL SDL files read using the
ReadandGlobtools. - Boundary markers: Not explicitly defined for the schemas themselves.
- Capability inventory: Restricted bash access (
rover), file system operations (Read,Write,Edit,Glob,Grep). - Sanitization: Data is processed by the formal
roverCLI parser during validation.
Audit Metadata