weekly-pipeline-digest
Fail
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: CRITICALPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from external HubSpot deal records to generate digests. Malicious text within deal names or descriptions could attempt to influence the agent's behavior or output.
- Ingestion points: Reads HubSpot deal stages, names, and metrics (SKILL.md).
- Boundary markers: None; there are no clear delimiters or instructions to the agent to treat the external HubSpot data as potentially untrusted.
- Capability inventory: The skill is restricted to read-only API calls for reporting and maintains internal snapshots; no dangerous system commands, file writes, or arbitrary network access patterns were detected.
- Sanitization: No sanitization or input validation logic is present to handle special characters or potential command strings within the HubSpot data.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata