release-notes
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill was evaluated for potential indirect prompt injection vulnerabilities because it processes external data such as pull request titles and commit logs. The skill includes robust mitigation by instructing the agent to strictly follow evidence-bound claims, separate internal noise from user-facing changes, and disclose any gaps in evidence. These constraints effectively prevent external data from overriding the agent's core instructions.
- [COMMAND_EXECUTION]: The skill's operation is restricted to reading local files and drafting text. It explicitly forbids dangerous operations such as creating Git tags, publishing hosted releases, or modifying package files. No execution of arbitrary shell commands or untrusted scripts was identified.
- [DATA_EXFILTRATION]: No network access or data exfiltration paths were found. The skill is prohibited from calling GitHub Releases APIs or performing any remote pushes, ensuring all processed data remains within the local environment.
- [CREDENTIALS_UNSAFE]: The skill does not contain hardcoded secrets or credentials. Its operations do not target sensitive locations like SSH keys or AWS configuration files, focusing exclusively on documentation and release artifacts.
Audit Metadata