Sweep
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various command-line tools for repository analysis, including Git for history and branch management, and language-specific linters like vulture, staticcheck, and cargo clippy. These operations are restricted by safety boundaries, requiring user approval for deletions and excluding sensitive directories like .git and node_modules.
- [EXTERNAL_DOWNLOADS]: The skill uses npx to run JavaScript analysis utilities such as knip and depcheck, which may fetch packages from the public npm registry. It also interacts with package managers for Python, Go, and Rust to audit dependencies, which is consistent with the skill's primary purpose of project maintenance.
- [PROMPT_INJECTION]: The skill processes untrusted local data (source code and configuration files) which presents a surface for indirect prompt injection.
- Ingestion points: Local file contents are read during the SCAN and ANALYZE steps (referenced in SKILL.md and maintenance-workflow.md).
- Boundary markers: No specific delimiters or "ignore instructions" warnings are defined for the content of analyzed files.
- Capability inventory: The skill can execute shell commands via npx and delete files (governed by user confirmation rules).
- Sanitization: No explicit sanitization or filtering of file content is described before processing by analysis tools.
Audit Metadata