safe-ai-essencelisted
Install: claude install-skill TeddyPika/Essence-Extract
# Safe AI Essence
Summarize the safety and governance posture of an AI system from the material provided. Focus on the actual safeguards, intended use, risk areas, and remaining gaps; do not present speculative harms as if they were documented facts.
## Process
1. Identify the AI system, model, workflow, or policy being assessed and the context in which it is used.
2. Extract the stated safety goals, risk categories, intended boundaries, and prohibited or discouraged uses.
3. Separate implemented controls from assumptions or future plans. Look for guardrails such as human review, content filtering, data minimization, evaluation, logging, red-teaming, and escalation paths.
4. Check whether the material describes monitoring, testing, incident response, or model update processes. Note the evidence for each claim and whether it is operational or merely aspirational.
5. Summarize residual risk and missing evidence. If the source does not support a conclusion, say so and identify what additional information would help.
## Response
Use a concise structure suitable to the request:
- **AI safety essence:** the core safety message in one or two sentences.
- **System scope and intended use:** what the model or workflow is meant to do and where it is not meant to be used.
- **Risks and harms:** the key categories of risk or harm described by the source material.
- **Controls and mitigations:** the concrete safeguards, evaluation methods, and governance practices in place.
- **Resid