changelog
Pass
Audited by Gen Agent Trust Hub on Jul 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill outlines a legitimate development process for maintaining changelogs. It utilizes a specialized CLI (
agent-native) and internal packages (@agent-native/core) to automate the creation and merging of markdown files. - [COMMAND_EXECUTION]: The skill uses the
agent-nativecommand to add and release changelog entries. These commands are purpose-built for the changelog workflow and do not exhibit signs of arbitrary or malicious command execution. - [INDIRECT_PROMPT_INJECTION]: The skill describes a process where the agent summarizes changes and writes them to files. While this represents a data ingestion surface, the risk is negligible within the context of generating standard documentation.
- Ingestion points: Descriptions provided to
agent-native changelog addinSKILL.md. - Boundary markers: None explicitly required for the CLI arguments.
- Capability inventory: File system writes (
changelog/*.md), modifications (CHANGELOG.md), and deletions (cleanup of pending files) as documented inSKILL.md. - Sanitization: Not explicitly mentioned, as the data is intended for static documentation files.
Audit Metadata