report-editor
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes and modifies RDLX-JSON reports and manifest metadata from external sources, which is an untrusted data ingestion surface.
- Ingestion points: Files are downloaded from a remote platform branch using
dxs report downloadand then inspected to understand their layout and data bindings. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the agent when processing the content of existing reports, such as field labels or expressions.
- Capability inventory: The skill possesses capabilities to read files, perform layout modifications (move, set, remove, batch), modify datasets, and upload the finalized components back to the remote platform via
dxs report upload. - Sanitization: No specific sanitization or validation of the report JSON content or expressions is defined before the agent processes the data or re-uploads it.
Audit Metadata