ds-content-perf
Warn
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: MEDIUMDYNAMIC_CONTEXT_INJECTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The skill uses the
!command`` syntax (specificallycat .agents/product-marketing-context.md) to execute a shell command and inject its output into the context at load time. This allows the skill to automatically read local files without user interaction.- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external analytics sources (Google Analytics 4 and Google Search Console) and user-provided files (CSV, TSV, JSON), which creates an attack surface for indirect prompt injection. - Ingestion points: GA4 API data, Search Console API data, manual data uploads, and the
.agents/product-marketing-context.mdfile. - Boundary markers: None identified; the skill does not use specific delimiters or instructions to ignore potential commands within the ingested data.
- Capability inventory: The skill has the ability to execute shell commands and Python scripts via the
Bashtool. - Sanitization: While the skill uses
ds_utils.pyto process and validate data formats, it does not perform sanitization to prevent data content from being interpreted as instructions by the LLM.- [COMMAND_EXECUTION]: The skill invokes theBashtool to run a local Python script (ds_utils.py) for data processing, validation, and content classification tasks.
Audit Metadata