optise-helix-eu-trust-centre
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a purely instructional template engine. It does not include executable code, scripts, or external dependencies. All logic is contained within Markdown and YAML instructions.
- [PROMPT_INJECTION]: No evidence of malicious prompt injection. The instructions include 'MANDATORY' and 'CRITICAL' markers, but these are used for internal control flow (e.g., Session Hygiene, URL verification) and safety adherence, rather than bypassing model guardrails.
- [EXTERNAL_DOWNLOADS]: The skill uses
web_fetchandweb_searchtools to verify user-provided URLs and factual claims (e.g., GDPR status, certifications). These operations are restricted to legitimate business verification and are governed by a '4-tier source hierarchy' that prioritizes official and reputable sources. - [DATA_EXFILTRATION]: No exfiltration patterns detected. The skill specifically instructs the agent to use placeholders (
[User to add: ...]) for sensitive information like data residency or DPO contacts if the user has not explicitly provided them, preventing the accidental leakage or fabrication of sensitive compliance data. - [INDIRECT_PROMPT_INJECTION]: The skill contains a 'Section 0' rule to treat all fetched or pasted content as untrusted data and explicitly commands the agent never to follow instructions contained within that data. It uses strict boundary markers and a verification log to maintain context integrity.
Audit Metadata