claim-chart
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for legal professionals to map evidence to patent claims or civil legal elements. Its functionality is consistent with its stated purpose.
- [DATA_EXPOSURE]: The skill accesses local configuration files (
~/.claude/plugins/config/...) and workspace-specific files (matter.md,history.md). This behavior is standard for an agent-based tool managing persistent case data and does not involve exfiltration to external domains. - [SAFE]: The skill explicitly instructs the agent to escape characters that could trigger formula execution in spreadsheet software (such as
=,+,-,@) by adding a single quote prefix. This is a proactive security measure against CSV/Excel Formula Injection attacks. - [PROMPT_INJECTION]: No malicious prompt injection patterns, role-play bypasses, or instructions to ignore safety guidelines were detected in the skill instructions or templates.
- [COMMAND_EXECUTION]: The skill does not contain any shell command execution (
!command) or subprocess calls. All operations are descriptive instructions for the agent to follow within its standard operational bounds.
Audit Metadata