check-readiness
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub issue bodies, comments, and implementation plans to identify requirements and define readiness criteria. This content is used to guide the agent's workflow and subsequent actions. An attacker could embed malicious instructions in these comments to influence the agent's behavior during verification or merging.
- Ingestion points: Data retrieved via
gh issue view(title, body, comments) as described in theSKILL.mdworkflow. - Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore instructions embedded within the processed issue content.
- Capability inventory: The skill can read repository files, spawn subagents, execute CLI tools (
gh,coderabbit), and invoke other skills such as$autofixandmerge-it. - Sanitization: No explicit sanitization or validation of the natural language content extracted from GitHub is defined in the workflow.
- [COMMAND_EXECUTION]: The skill utilizes command-line tools to interact with the repository and external platforms.
- Evidence: The workflow involves executing
gh issue viewfor data retrieval andcoderabbit reviewfor automated code analysis. - [EXTERNAL_DOWNLOADS]: The skill integrates with CodeRabbit, a well-known third-party automated code review service.
- Evidence:
SKILL.mddescribes invoking the$coderabbit-cliskill and runningcoderabbit reviewcommands to facilitate readiness checks. - Note: The skill includes a safety requirement to disclose that repository context may be uploaded to an external service and to notify the user before running the tool.
Audit Metadata