printing-press-publish
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and downloads tools from established sources. It recommends installing the
cli-printing-presstool via Go from the vendor's GitHub repository and utilizesgovulncheckfrom the official Go repository (golang.org) for security validation. - [COMMAND_EXECUTION]: The skill makes extensive use of system commands including
gitfor repository management,ghfor GitHub API interactions, andgofor building and validating code. These operations are necessary for the skill's primary function of packaging and publishing software. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from local CLI project directories, including manifests and documentation.
- Ingestion points: Reads
.printing-press.json,README.md, and project source files (SKILL.md) from the target CLI directory. - Boundary markers: Relies on structured validation checks performed by the
cli-printing-press validatetool and user confirmation prompts before performing sensitive actions like replacing existing PRs. - Capability inventory: Includes the ability to execute shell commands, perform Go builds, modify files, and interact with remote Git repositories and the GitHub API.
- Sanitization: Implements a comprehensive "Secret & PII Protection" layer, which includes mandatory binary scanning for vendor-prefix tokens (e.g., API keys), exact-value scans for known secrets, and a two-tier PII pattern scanning process to redact sensitive information before publication.
- [DATA_EXFILTRATION]: While the skill interacts with the network to push code to GitHub, it contains explicit defensive mechanisms to prevent the exfiltration of credentials or personally identifiable information (PII) to the public library repository.
Audit Metadata