secure-development-lifecycle
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a procedural guide for software security and does not include any executable scripts, shell commands, or network exfiltration logic.
- [EXTERNAL_DOWNLOADS]: The skill references several external resources, including security policies on GitHub and reference implementations at domains like
blacktrigram.comandciacompliancemanager.com. These links are for documentation and informational purposes, aligning with the vendor's ecosystem. - [PROMPT_INJECTION]: The skill includes quality-control instructions such as the "AI FIRST principle," which directs the agent to iterate on its output to ensure quality. These are benign instructional markers rather than attempts to bypass safety filters or ignore system constraints.
- [CREDENTIALS_UNSAFE]: The skill explicitly mandates "Zero Hard-Coded Credentials" and provides guidance on using secure secret management systems like GitHub Secrets, demonstrating a strong security posture.
Audit Metadata