coderabbit-cost-tuning
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for cost optimization of the CodeRabbit service. It provides guidance on seat management and configuration, along with a bash script for auditing review coverage.
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external data by querying the GitHub API for pull request details and comments.
- Ingestion points: Pull request numbers and comment metadata are fetched in
SKILL.md(Step 6) using thegh apicommand. - Boundary markers: Data is processed using
jqto extract specific fields, but no explicit 'ignore instructions' delimiters are used for the raw API output. - Capability inventory: The skill utilizes
Bash(gh:*)to perform API queries and calculate metrics. - Sanitization: Input is constrained by
jqfiltering, which extracts a count (length) rather than text content, minimizing the risk of the agent interpreting malicious data as instructions.
Audit Metadata