content-decay
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external Search Console page exports. Although the processing logic in
resources/content-decay.jsis safe and limited to mathematical analysis, the ingested data is returned to the agent's context, which is a standard surface for indirect prompt injection if the source data (e.g., page titles or URLs) is compromised.\n - Ingestion points: The skill takes a JSON payload representing Search Console data as input to the
content-decay.jsscript via the command line.\n - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the agent when processing the external data.\n
- Capability inventory: The analysis script is restricted to local processing and lacks tools for network communication, file system modifications (beyond reading its own help text), or shell command execution.\n
- Sanitization: The script performs numeric validation but does not sanitize or escape the string content of page identifiers before they are rendered in the final report.
Audit Metadata