dspy-program-security
CommunitySecure your DSPy prompts, datasets, and tools.
System Documentation
What problem does it solve?
DSPy compiles your examples into prompts; if optimizer datasets or tool wrappers contain unsafe content, security gaps propagate into production prompts. This skill provides a structured review to identify and remediate unsafe signatures, data leakage from optimizer training, boundary violations between prompt programs and tooling, and enforcement of strict output typing.
Core Features & Use Cases
- Review module signatures, adapters, and prompt-program composition to ensure isolation and safe tool usage.
- Enforce strict typed outputs and per-module tool allowlists to limit surface area for risk.
- Establish a dataset curation pipeline to prevent insecure optimizer data from poisoning compiled prompts and deployed artifacts.
- Perform a compiled-program review before deployment to detect injected content or misconfigurations.
Quick Start
Review a DSPy program's security posture by inspecting optimizer datasets, compiled prompts, and tool wrappers, and implement a strict review workflow.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: dspy-program-security Download link: https://github.com/maruakshay/mii-ai-security/archive/main.zip#dspy-program-security Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.