Scan prompts for PII and semantic issues, returning JSON findings.
npx skills add https://github.com/CraigHutchinson/Prism --skill prism-sanitize
Please help me install this Agent Skill. Skill: prism-sanitize Source: https://github.com/CraigHutchinson/Prism/tree/main/.claude/skills/prism-sanitize Command: npx skills add https://github.com/CraigHutchinson/Prism --skill prism-sanitize
This Skill provides internal sanitization analysis to enhance prompt security and compliance, preventing data leaks and improving prompt quality.
To begin sanitization analysis, execute the following commands: python scripts/pii_scan.py --json "<prompt>" and reference sanitization rules with cat .cursor/skills/prism/sanitization-rules.md.
High-intent search queries and answers about installing and using this skill.
Prompt sanitization prevents PII data leaks by scanning prompts for personal information and semantic issues like indirect injections. This process uses Python scripts to detect security vulnerabilities and outputs findings in structured JSON format for compliance review.
PII detection in prompt engineering identifies Personally Identifiable Information within text inputs. Semantic analysis enhances this by checking for ambiguous authority statements and indirect injections, ensuring comprehensive security and compliance before prompt execution.
Yes, you need a Python environment to execute prompt sanitization and PII scanning scripts. The advanced analysis requires specific Python scripts to process prompts, detect personal information leaks, and generate structured JSON outputs for security compliance.
Yes, you can detect indirect injections and ambiguous authority statements automatically through semantic analysis. The sanitization process evaluates prompt content for these semantic security issues alongside PII detection, returning all findings in structured JSON format.
The best way to format prompt security analysis results is using structured JSON outputs. This format captures PII detection findings and semantic analysis results, enabling seamless downstream processing and integration with automated data leak prevention workflows.
Prompt sanitization outputs structured JSON to facilitate automated downstream processing of security findings. JSON formatting allows systems to programmatically parse detected PII, semantic issues, and security vulnerabilities, integrating smoothly into existing data leak prevention pipelines.
GeeksfinoDiagnose OpenSkills runtime failures in Landlock, seatbelt, and WASM sandboxing.
atomicinnovationDefine a standardized JSON output format for plan review agents.
4lermanAudit repositories for sensitive data and manage environment variables securely.
framerslabDetect and redact personally identifiable information from text using regex, NLP, and NER.
