# Goochbeater/Spiritual-Spell-Red-Teaming

A repo for jailbreaking various LLMs, mainly Claude

Repository: https://github.com/Goochbeater/Spiritual-Spell-Red-Teaming
Canonical: https://ross.abutalabs.com/products/spiritual-spell-red-teaming
License Family: other
Last push: 2026-08-10T00:45:55+00:00

## Health v2 (maintenance only)
Score: 59/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 35, longevity 18
- inputs: {"age_days": 261, "days_push": 24, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3313, forks 639 (observed 2026-08-28T04:07:55.889207+00:00)

## What it is
A curated collection of jailbreak prompts and red-teaming techniques targeting large language models, primarily Claude. It serves as an adversarial testing resource for probing LLM safety guardrails and refusal behaviors.

## Use cases
- test whether an LLM refuses harmful requests
- red-team Claude's safety guardrails
- study jailbreak prompt patterns against LLMs
- evaluate robustness of AI safety filters
- collect examples of adversarial prompts for research

## When to choose
- you are doing authorized red-teaming or safety evaluation of LLMs
- you research LLM jailbreak techniques and refusal bypasses
- you need a reference corpus of adversarial prompts for defensive analysis

## When to avoid
- you want a production tool or library with an API
- you need licensed, maintained software - the repo has no license
- you intend to use it to bypass safety measures on systems you are not authorized to test

## Facets
- artifact type: dataset
- maturity: active
- function: prompt-engineering, security, llm-inference
- domain: artificial-intelligence, large-language-models, security, penetration-testing
- platform: cross-platform
- tags: jailbreak-prompts, red-teaming, llm-safety, adversarial-prompts, claude, prompt-collection

## Member repositories
- Goochbeater/Spiritual-Spell-Red-Teaming (main) score 59

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:55.889207+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:42:22.220054+00:00, confidence not recorded.
  - readme: https://github.com/Goochbeater/Spiritual-Spell-Red-Teaming (fetched 2026-08-28T04:07:55.889207+00:00, sha 2472f9f9105b)
- Data as of 2026-08-30T08:39:29.467469+00:00.