coverage-diff-reporter
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill documentation describes the use of standard shell commands within CI/CD workflows, such as
npm test,git checkout, and executing local Python scripts (e.g.,python scripts/coverage_diff.py). These operations are consistent with the skill's primary function of processing local coverage data. - [EXTERNAL_DOWNLOADS]: The skill utilizes well-known GitHub Actions to perform its tasks. It fetches baseline artifacts using
dawidd6/action-download-artifact@v3and posts results usingmarocchino/sticky-pull-request-comment@v2. These are standard tools for DevOps automation. - [INDIRECT_PROMPT_INJECTION]: The skill has a potential attack surface for indirect prompt injection as it ingests and processes untrusted data from coverage files to generate reports.
- Ingestion points: The skill reads coverage data from files such as
coverage/lcov.infoandbaseline/lcov.info(referenced inSKILL.md). - Boundary markers: No explicit boundary markers or warnings for the agent to ignore embedded instructions are present in the provided snippets.
- Capability inventory: The skill performs file reads, executes local scripts, and writes output to markdown files which are then posted as PR comments.
- Sanitization: The provided Python code snippets show data extraction and arithmetic for coverage deltas but do not include explicit sanitization of input strings (like file paths) before they are interpolated into the markdown report.
Audit Metadata