working-on-github-issue
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes GitHub CLI (
gh) commands to retrieve issue metadata (gh issue view) and update issue assignments (gh issue edit --add-assignee @me). It also updates GitHub Project items based on the provided stage. - [EXTERNAL_DOWNLOADS]: The skill provides guidance to install the
new-branchdependency from the vendor's repository (patinaproject/skills) usingnpx skills@latest. The suggested command correctly includes thenpm_config_ignore_scripts=trueflag to prevent the execution of potentially malicious lifecycle scripts during the installation process. - [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it ingests untrusted data from GitHub issues.
- Ingestion points: Data is ingested from issue references, URLs, and metadata fields (assignees, project items) using the
ghCLI as defined in SKILL.md. - Boundary markers: No specific delimiters or instructions are provided to the agent to distinguish between the ingested issue data and its core operating instructions.
- Capability inventory: The skill can execute the
new-branchchild skill and modify issue assignments and project statuses via theghtool. - Sanitization: The instructions do not include any verification or sanitization steps for the content retrieved from external GitHub issues before it is processed by the agent.
Audit Metadata