SkillForge

Appearance

Get Started

AI/ML · Global library

AI Safety Evaluator

Design and execute comprehensive safety evaluations for AI systems with red-teaming, adversarial testing, and safety metric frameworks

CodexClaude CodeKimi Codeorchestrator-mcp
Open pack on GitHubBack to catalog

Best use case

Use AI Safety Evaluator when you need to design and execute comprehensive safety evaluations for AI systems with red-teaming, adversarial testing, and safety metric frameworks, especially when the work is driven by safety evaluation and red team.

Trigger signals

safety evaluationred teamadversarial testsafety metricsharmful contentjailbreak

Validation hooks

coverage-checkthreshold-validation

Install surface

Copy the exact command path you need.

Inspect

pip install "orchestrator-mcp[dashboard]"
orchestrator-mcp skills show ai-safety-evaluator

Use

orchestrator-mcp skills export ai-safety-evaluator --to ./skillforge-packs
# copy the exported pack into your preferred agent environment

Export

cp -R skills/ai-safety-evaluator ./your-agent-skills/ai-safety-evaluator
# or open skills/ai-safety-evaluator/SKILL.md in a markdown-first client

File patterns

*.pyeval*.pysafety/*.pytest*.py

Model preferences

claude-opus-4gpt-4oclaude-haiku-3

Related skills

Adjacent packs to compose next.

AI/MLGlobal library

Agent Lifecycle Manager

Open pack

Manage complete agent lifecycles from initialization through graceful shutdown with health monitoring, scaling, and resource optimization

CodexClaude Code
AI/MLGlobal library

Agent Memory Designer

Open pack

Design short-term, long-term, and episodic memory layers for agents without turning retrieval into an unbounded context leak.

CodexClaude Code
Apache-2.0 licensedUse it, fork it, redistribute it. No strings attached.Safety-first by designSecret redaction, release scanning, and no shell execution. Constraints are public.GitHub nativeEvery pack is a real folder. Fork, edit, PR. No hidden backend or proprietary pipeline.