Index

A

Ablation_Dan_11_0 (class in garak.probes.dan) accuracy (MustRefuteClaimModel attribute) active (Ablation_Dan_11_0 attribute) (AgentBreaker attribute) (AnsiEscaped attribute) (AnsiRaw attribute) (AnsiRawTokenizerHF attribute) (Anthropomorphisation attribute) (AntiDAN attribute) (Attack_On_Author attribute) (Attack_On_Commenter attribute) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (AzureOpenAIGenerator attribute) (BadCharacters attribute) (BEAST attribute) (BedrockGenerator attribute) (Blank attribute) (Buff attribute) (Bullying attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteKey attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlurs attribute) (ContinueSlursReclaimedSlursFull attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWild attribute) (DanInTheWildFull attribute) (Dart attribute) (Davidjl attribute) (Deadnaming attribute) (Detector attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (DRA attribute) (DRAAdvanced attribute) (DUDE attribute) (EICAR attribute) (Evasion attribute) (FactSnippetMixin attribute) (FalseAssertion attribute) (FigStep attribute) (FigStepFull attribute) (FITD attribute) (Flirtation attribute) (FunctionMasking attribute) (FutureTense attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (Generator attribute) (GetKey attribute) (Glitch attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivity attribute) (GraphConnectivityFull attribute) (GroqChat attribute) (GTphish attribute) (GTUBE attribute) (Harness attribute) (HF_Files attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HomoglyphObfuscation attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (InformationHazard attribute) (InjectAscii85 attribute) (InjectAtbash attribute) (InjectBase16 attribute) (InjectBase2048 attribute) (InjectBase32 attribute) (InjectBase64 attribute) (InjectBraille attribute) (InjectEcoji attribute) (InjectHex attribute) (InjectMime attribute) (InjectMorse attribute) (InjectNato attribute) (InjectQP attribute) (InjectROT13 attribute) (InjectSneakyBits attribute) (InjectUnicodeTagChars attribute) (InjectUnicodeVariantSelectors attribute) (InjectUU attribute) (InjectZalgo attribute) (Insult attribute) (Insult_Experimental attribute) (Jailbreak attribute) (JavaScript attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (LatentWhoisSnippetFull attribute) (Likely_To_Reject attribute) (LLMGenerator attribute) (MaliciousUses attribute) (MarkdownImageExfil attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (MisinformationHarms attribute) (MLM attribute) (ModelAsJudge attribute) (MustContradictNLI attribute) (NonFullMixin attribute), [1] (NVOpenAIChat attribute) (Obscene attribute) (OllamaGenerator attribute) (OpenAICompatible attribute) (OpenAIGenerator attribute) (PackageHallucinationDetector attribute) (PAIR attribute) (PastTense attribute) (PastTenseFull attribute) (Payload attribute) (Perl attribute) (PIILeakQuadruplet attribute) (PIILeakTriplet attribute) (PIILeakTwin attribute) (PIILeakUnstructured attribute) (PlaygroundMarkdownExfil attribute) (Primes attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute), [1] (Profanity_Experimental attribute) (Python attribute) (QuackMedicine attribute) (RakuLand attribute) (Refusal attribute) (Repeat attribute) (RepeatedToken attribute), [1] (RepeatExtended attribute) (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (Ruby attribute) (Rust attribute) (Senators attribute) (SenatorsFull attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (SexualContent attribute) (Sexualisation attribute) (Sexually_Explicit attribute) (Slurs attribute) (SlurUsage attribute) (Spam attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (STAN attribute) (StringAssemblyDataExfil attribute) (SubFunctions attribute) (Substances attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (Threat attribute) (Threat_Experimental attribute) (ThreatenJSON attribute) (TopLevel attribute) (Tox attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (TranslationMixin attribute) (Unsubstantial attribute) (WhoIsRiley attribute) (Win10 attribute) (WordnetAllowedWords attribute) (WordnetBlockedWords attribute) (WordnetControversial attribute)

affiliations (PIILeakUnstructured attribute) AgentBreaker (class in garak.probes.agent_breaker) AgentBreakerResult (class in garak.detectors.agent_breaker) aliases (ContinueSlursReclaimedSlurs attribute) (EICAR attribute) (GCG attribute) (GCGCached attribute) (GTphish attribute) (GTUBE attribute) (Repeat attribute) (RepeatExtended attribute) all_outputs (Attempt property) AnsiEscaped (class in garak.probes.ansiescape) AnsiRaw (class in garak.probes.ansiescape) AnsiRawTokenizerHF (class in garak.probes.ansiescape) Anthro (class in garak.detectors.lmrc) AnthropicGenerator (class in garak.generators.anthropic) Anthropomorphisation (class in garak.probes.lmrc) AntiDAN (class in garak.detectors.dan) (class in garak.probes.dan) AnyCode (class in garak.detectors.malwaregen) AnyNonspaceOutput (class in garak.detectors.any) AnyOutput (class in garak.detectors.any) api_version (AzureOpenAIGenerator attribute) ApiKey (class in garak.detectors.apikey) APIKeyMissingError append_report_object() (in module garak.analyze.report_digest) args (TransientConfig attribute) as_dict() (Attempt method) atbash() (InjectAtbash static method) Attack_On_Author (class in garak.detectors.perspective) Attack_On_Commenter (class in garak.detectors.perspective) AttackRogueString (class in garak.detectors.promptinject) AttackState (class in garak.probes.agent_breaker) Attempt (class in garak.attempt) attempts_history (AttackState attribute) AudioAchillesHeel (class in garak.probes.audio) AutoDAN (class in garak.probes.dan) AutoDANCached (class in garak.probes.dan) AzureOpenAIGenerator (class in garak.generators.azure)

B

BadCharacters (class in garak.probes.badchars) BadGeneratorException banner() (in module garak.interactive) Base64 (class in garak.buffs.encoding) base_instructions (TranslationMixin attribute) base_prompt (Payload attribute) base_prompts (Evasion attribute) (SubFunctions attribute) (TopLevel attribute) BEAST (class in garak.probes.suffix) BedrockGenerator (class in garak.generators.bedrock) Blank (class in garak.generators.test) (class in garak.probes.test)

BlankVision (class in garak.generators.test) braille() (in module garak.probes.encoding) Buff (class in garak.buffs.base) buff() (Buff method) BuffManager (class in garak._config) buffs (BuffManager attribute) build_digest() (in module garak.analyze.report_digest) build_html() (in module garak.analyze.report_digest) Bullying (class in garak.probes.lmrc) Bypass (class in garak.probes.doctor) BypassLeet (class in garak.probes.doctor)

C

cache_dir (TransientConfig attribute) Calibration (class in garak.analyze.calibration) case_sensitive (PIILeak attribute) CharCode (class in garak.buffs.encoding) ChatGPT_Developer_Mode_RANTI (class in garak.probes.dan) ChatGPT_Developer_Mode_v2 (class in garak.probes.dan) ChatGPT_Image_Markdown (class in garak.probes.dan) clear_history() (Generator method) ClozeProbeMixin (class in garak.probes.leakreplay) cmd2_default_help_category (GarakCommands attribute) CohereGenerator (class in garak.generators.cohere) ColabAIDataLeakage (class in garak.probes.web_injection) COMPETE_WITH_SOTA (Tier attribute) CompleteKey (class in garak.probes.apikey) CompleteProbeMixin (class in garak.probes.leakreplay) config_dir (TransientConfig attribute)

config_hash (LangChainServeLLMGenerator attribute) ConfigFailure constructor() (in module garak.detectors.perspective) (in module garak.probes.promptinject) content (Turn attribute) contexts (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (TranslationMixin attribute) Continuation (class in garak.detectors.continuation) ContinueSlursReclaimedSlurs (class in garak.probes.continuation) ContinueSlursReclaimedSlursFull (class in garak.probes.continuation) Conversation (class in garak.attempt) current_attack_prompt (AttackState attribute) current_target (AttackState attribute) current_tool_analysis (AttackState attribute)

D

DAN (class in garak.detectors.dan) Dan_10_0 (class in garak.probes.dan) Dan_11_0 (class in garak.probes.dan) Dan_6_0 (class in garak.probes.dan) Dan_6_2 (class in garak.probes.dan) Dan_7_0 (class in garak.probes.dan) Dan_8_0 (class in garak.probes.dan) Dan_9_0 (class in garak.probes.dan) DAN_Jailbreak (class in garak.probes.dan) DANDetectorMeta (class in garak.detectors.dan) DanInTheWild (class in garak.probes.dan) DanInTheWildFull (class in garak.probes.dan) DANJailbreak (class in garak.detectors.dan) DANProbeMeta (class in garak.probes.dan) Dart (class in garak.detectors.packagehallucination) (class in garak.probes.packagehallucination) data (Message property) data_checksum (Message attribute) data_dir (TransientConfig attribute) data_path (Message attribute) data_type (Message attribute) Davidjl (class in garak.probes.goodside) Deadnaming (class in garak.probes.lmrc) DecodeApprox (class in garak.detectors.encoding) DecodeMatch (class in garak.detectors.encoding) default() (GarakTerminal method) (PluginEncoder method) deprecation_notice() (in module garak.command) description (Attack_On_Author attribute) (Attack_On_Commenter attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (Flirtation attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HumanChatbox attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (InformationHazard attribute) (Insult attribute) (Insult_Experimental attribute) (Jailbreak attribute) (Likely_To_Reject attribute) (MaliciousUses attribute) (MisinformationHarms attribute) (ModelAsJudge attribute) (MustRefuteClaimModel attribute) (Obscene attribute) (Profanity attribute) (Profanity_Experimental attribute) (Refusal attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (Sexually_Explicit attribute) (Spam attribute) (Threat attribute) (Threat_Experimental attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) detect() (AgentBreakerResult method) (AnyCode method) (AnyNonspaceOutput method) (AnyOutput method) (ApiKey method) (Attack_On_Author method) (Attack_On_Commenter method) (AttackRogueString method) (Continuation method) (DecodeApprox method) (DecodeMatch method) (Detector method) (DontStartNo method) (DontStartYes method) (Down method) (Fail method) (FileDetector method) (Flirtation method) (Glitch method) (HFDetector method) (Identity_Attack method) (Identity_Attack_Experimental method) (in module garak.detectors.perspective) (Incoherent method) (Inflammatory method) (Insult method) (Insult_Experimental method) (Jailbreak method) (JailbreakOnlyAdversarial method) (JinjaTemplateInjectionDetector method) (Likely_To_Reject method) (MarkdownExfilBasic method) (MarkdownExfilContent method) (MarkdownExfilExtendedImage method) (MarkdownExfilExtendedNonImage method) (MitigationBypass method) (ModelAsJudge method) (ModernBERTRefusal method) (MustContradictNLI method) (MustRefuteClaimKW method) (MustRefuteClaimModel method) (Obscene method) (PackageHallucinationDetector method) (Pass method) (Passthru method) (PIILeak method) (PlainJSON method) (Prefixes method) (Profanity method) (Profanity_Experimental method) (PromptExtraction method) (PromptExtractionVerbatim method) (PythonCodeExecution method) (Random method) (Refusal method) (RefusalOnlyAdversarial method) (RepeatDiverges method) (RepeatedToken method) (Severe_Toxicity method) (Severe_Toxicity_Experimental method) (Sexually_Explicit method) (Spam method) (SQLiEcho method) (StartsWith method) (StringDetector method) (Threat method) (Threat_Experimental method) (Toxicity method) (Toxicity_Experimental method) (TriggerListDetector method) (Unsubstantial method) (Up method) (Win5x5 method)

Detector (class in garak.detectors.base) DevMode (class in garak.detectors.dan) DID_ENV_VAR (WatsonXGenerator attribute) Director (class in garak.payloads) DiscriminationExclusionToxicityHatefulOffensive (class in garak.probes.donotanswer) dna_category (DiscriminationExclusionToxicityHatefulOffensive attribute) (HumanChatbox attribute) (InformationHazard attribute) (MaliciousUses attribute) (MisinformationHarms attribute) do_list() (GarakCommands method) do_probe() (GarakCommands method) do_quit() (GarakTerminal method) doc_uri (AgentBreaker attribute) (AgentBreakerResult attribute) (AnsiEscaped attribute) (AnsiRawTokenizerHF attribute) (Anthropomorphisation attribute) (AnyOutput attribute) (AttackRogueString attribute) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (BadCharacters attribute) (BEAST attribute) (Blank attribute) (Buff attribute) (Bullying attribute) (Bypass attribute) (BypassLeet attribute) (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlursFull attribute) (DanInTheWildFull attribute) (Davidjl attribute) (Deadnaming attribute) (Detector attribute) (DRA attribute) (EICAR attribute) (EncodingMixin attribute) (Escaped attribute) (FactSnippetMixin attribute) (FalseAssertion attribute) (Fast attribute) (FigStepFull attribute) (FITD attribute) (FunctionMasking attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (Glitch attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivityFull attribute) (GTphish attribute) (GTUBE attribute) (HomoglyphObfuscation attribute) (HypotheticalResponse attribute) (InjectAtbash attribute) (InjectBraille attribute) (InjectEcoji attribute) (InjectLeet attribute) (InjectMorse attribute) (InjectNato attribute) (InjectSneakyBits attribute) (InjectUnicodeTagChars attribute) (InjectUnicodeVariantSelectors attribute) (InjectZalgo attribute) (Jailbreak attribute) (JinjaTemplateInjectionDetector attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (LatentWhoisSnippetFull attribute) (LRLBuff attribute) (MarkdownExfilBasic attribute) (MarkdownExfilContent attribute) (MarkdownExfilExtendedImage attribute) (MarkdownExfilExtendedNonImage attribute) (MarkdownImageExfil attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (Martin_haDistilbert attribute) (MitigationBypass attribute) (MLM attribute) (ModernBERTRefusal attribute) (PackageHallucinationProbe attribute) (PAIR attribute) (PastTenseFull attribute) (PegasusT5 attribute) (PIILeak attribute) (PIILeakExact attribute) (PIILeakProbeMixin attribute) (PlainJSON attribute) (PlaygroundMarkdownExfil attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute) (Puppetry attribute) (PythonCodeExecution attribute) (QuackMedicine attribute) (Raw attribute) (Repeat attribute) (RepeatedToken attribute), [1] (RTPBlank attribute) (S_nlpDetox attribute) (SenatorsFull attribute) (SexualContent attribute) (Sexualisation attribute) (Single attribute) (SlursReclaimedSlurs attribute) (SlurUsage attribute) (SQLiEcho attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (SQLiSuccess attribute) (StringAssemblyDataExfil attribute) (Substances attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (ThreatenJSON attribute) (Tox attribute) (TranslationMixin attribute) (WhoIsRiley attribute) (Win10 attribute) DontStartNo (class in garak.detectors.snowball) DontStartYes (class in garak.detectors.snowball) Down (class in garak.detectors.shields) DRA (class in garak.probes.dra) DRAAdvanced (class in garak.probes.dra) DUDE (class in garak.detectors.dan) (class in garak.probes.dan)

E

EICAR (class in garak.detectors.knownbadsignatures) (class in garak.probes.av_spam_scanning) enabled() (in module garak.langservice) encoding_funcs (InjectAscii85 attribute) (InjectAtbash attribute) (InjectBase16 attribute) (InjectBase32 attribute) (InjectBase64 attribute) (InjectBraille attribute) (InjectEcoji attribute) (InjectHex attribute) (InjectLeet attribute) (InjectMorse attribute) (InjectNato attribute) (InjectQP attribute) (InjectROT13 attribute) (InjectSneakyBits attribute) (InjectUnicodeVariantSelectors attribute) (InjectZalgo attribute) encoding_name (InjectAscii85 attribute) (InjectAtbash attribute) (InjectBase16 attribute) (InjectBase2048 attribute) (InjectBase32 attribute) (InjectBase64 attribute) (InjectBraille attribute) (InjectEcoji attribute) (InjectHex attribute) (InjectLeet attribute) (InjectMime attribute) (InjectMorse attribute) (InjectNato attribute) (InjectQP attribute) (InjectROT13 attribute) (InjectSneakyBits attribute) (InjectUnicodeTagChars attribute) (InjectUnicodeVariantSelectors attribute) (InjectUU attribute) (InjectZalgo attribute) EncodingMixin (class in garak.probes.encoding) encodings (MarkdownImageExfil attribute) end_run() (in module garak.command) endpoint (Attack_On_Author attribute) (Attack_On_Commenter attribute) (Flirtation attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (Insult attribute) (Insult_Experimental attribute) (Likely_To_Reject attribute) (Obscene attribute) (Profanity attribute) (Profanity_Experimental attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (Sexually_Explicit attribute) (Spam attribute) (Threat attribute) (Threat_Experimental attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) ENDPOINT_ENV_VAR (AzureOpenAIGenerator attribute) enumerate_plugins() (in module garak._plugins) ENV_VAR (AnthropicGenerator attribute) (Attack_On_Author attribute) (Attack_On_Commenter attribute) (AzureOpenAIGenerator attribute) (CohereGenerator attribute) (Flirtation attribute) (GroqChat attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (InferenceAPI attribute) (Inflammatory attribute) (Insult attribute) (Insult_Experimental attribute) (LangChainServeLLMGenerator attribute) (Likely_To_Reject attribute) (LRLBuff attribute) (MistralGenerator attribute) (NeMoGuardrailsServer attribute) (NvcfChat attribute) (NVOpenAIChat attribute) (Obscene attribute) (OpenAICompatible attribute) (OpenAIGenerator attribute) (Profanity attribute) (Profanity_Experimental attribute) (RasaRestGenerator attribute) (ReplicateGenerator attribute) (RestGenerator attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (Sexually_Explicit attribute) (Spam attribute) (Threat attribute) (Threat_Experimental attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) (WatsonXGenerator attribute) (WebSocketGenerator attribute)

Escaped (class in garak.detectors.ansiescape) evaluate() (Evaluator method) Evaluator (class in garak.evaluators.base) Evasion (class in garak.probes.malwaregen) exec_types (FileIsExecutable attribute) export() (Report method) extended_detectors (Ablation_Dan_11_0 attribute) (AntiDAN attribute) (AutoDAN attribute) (AutoDANCached attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWildFull attribute) (DRA attribute) (DUDE attribute) (EncodingMixin attribute) (FalseAssertion attribute) (HF_Files attribute) (JinjaTemplatePythonInjection attribute) (LatentJailbreakFull attribute) (MarkdownImageExfil attribute) (Probe attribute) (Profanity attribute) (Repeat attribute) (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (SexualContent attribute) (Sexualisation attribute) (Slurs attribute) (SlurUsage attribute) (STAN attribute) (Win10 attribute) extra_dependency_names (AnthropicGenerator attribute) (AudioAchillesHeel attribute) (BedrockGenerator attribute) (Buff attribute) (CohereGenerator attribute) (Detector attribute) (FileIsExecutable attribute) (Harness attribute) (LangChainLLMGenerator attribute) (LiteLLMGenerator attribute) (LLaVA attribute) (LLMGenerator attribute) (MistralGenerator attribute) (NeMoGuardrails attribute) (OllamaGenerator attribute) (ReplicateGenerator attribute) (WebSocketGenerator attribute)

F

f (GCGCached attribute) FactSnippetMixin (class in garak.probes.latentinjection) Fail (class in garak.detectors.always) FalseAssertion (class in garak.probes.misleading) Fast (class in garak.buffs.paraphrase) FigStep (class in garak.detectors.visual_jailbreak) (class in garak.probes.visual_jailbreak) FigStepFull (class in garak.probes.visual_jailbreak) FileDetector (class in garak.detectors.base) FileIsExecutable (class in garak.detectors.fileformats) FileIsPickled (class in garak.detectors.fileformats) FITD (class in garak.probes.fitd) Flirtation (class in garak.detectors.perspective) follow_prompt_cap (ClozeProbeMixin attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlurs attribute) (ContinueSlursReclaimedSlursFull attribute) (LatentInjectionMixin attribute) (NonFullMixin attribute), [1]

from_dict() (Conversation static method) (Turn class method) from_notes() (AttackState class method) from_openai() (Conversation class method) fullname (AnthropicGenerator attribute) (MistralGenerator attribute) FunctionMasking (class in garak.probes.smuggling) FutureTense (class in garak.probes.phrasing) FutureTenseFull (class in garak.probes.phrasing)

G

garak._config module garak._plugins module garak.analyze.calibration module garak.analyze.report_digest module garak.attempt module garak.buffs.base module garak.buffs.encoding module garak.buffs.low_resource_languages module garak.buffs.lowercase module garak.buffs.paraphrase module garak.cli module garak.command module garak.detectors.agent_breaker module garak.detectors.always module garak.detectors.ansiescape module garak.detectors.any module garak.detectors.apikey module garak.detectors.base module garak.detectors.continuation module garak.detectors.dan module garak.detectors.divergence module garak.detectors.encoding module garak.detectors.exploitation module garak.detectors.fileformats module garak.detectors.goodside module garak.detectors.judge module garak.detectors.knownbadsignatures module garak.detectors.leakreplay module garak.detectors.lmrc module garak.detectors.malwaregen module garak.detectors.misleading module garak.detectors.mitigation module garak.detectors.packagehallucination module garak.detectors.perspective module garak.detectors.productkey module garak.detectors.promptinject module garak.detectors.propile module garak.detectors.shields module garak.detectors.snowball module garak.detectors.sysprompt_extraction module garak.detectors.unsafe_content module garak.detectors.visual_jailbreak module garak.detectors.web_injection module garak.evaluators.base module garak.exception module garak.generators.anthropic module garak.generators.azure module garak.generators.base module garak.generators.bedrock module garak.generators.cohere module garak.generators.function module garak.generators.ggml module garak.generators.groq module garak.generators.guardrails module garak.generators.huggingface module garak.generators.langchain module garak.generators.langchain_serve module garak.generators.litellm module garak.generators.llm module garak.generators.mistral module garak.generators.nim module garak.generators.nvcf module garak.generators.ollama module garak.generators.openai module garak.generators.rasa module garak.generators.replicate module garak.generators.rest module garak.generators.test module garak.generators.watsonx module garak.generators.websocket module garak.harnesses.base module garak.harnesses.probewise module garak.harnesses.pxd module garak.interactive module garak.langservice module garak.payloads module garak.probes._tier module garak.probes.agent_breaker module garak.probes.ansiescape module garak.probes.apikey module garak.probes.atkgen module garak.probes.audio module garak.probes.av_spam_scanning module garak.probes.badchars module garak.probes.base module garak.probes.continuation module garak.probes.dan module garak.probes.divergence module garak.probes.doctor module garak.probes.donotanswer module garak.probes.dra module garak.probes.encoding module garak.probes.exploitation module garak.probes.fileformats module garak.probes.fitd module garak.probes.glitch module garak.probes.goat module garak.probes.goodside module garak.probes.grandma module garak.probes.latentinjection module garak.probes.leakreplay module garak.probes.lmrc module garak.probes.malwaregen module garak.probes.misleading module garak.probes.packagehallucination module garak.probes.phrasing module garak.probes.promptinject module garak.probes.propile module garak.probes.realtoxicityprompts module garak.probes.sata module garak.probes.smuggling module

garak.probes.snowball module garak.probes.suffix module garak.probes.sysprompt_extraction module garak.probes.tap module garak.probes.test module garak.probes.topic module garak.probes.visual_jailbreak module garak.probes.web_injection module garak.report module GarakCommands (class in garak.interactive) GarakException GarakSubConfig (class in garak._config) GarakTerminal (class in garak.interactive) GCG (class in garak.probes.suffix) GCGCached (class in garak.probes.suffix) generate() (Generator method) (LLaVA method) generate_next_turn_attempt() (FITD method) generate_prompts() (MarkdownExfilMixin method) Generator (class in garak.generators.base) generator_family_name (AnthropicGenerator attribute) (AzureOpenAIGenerator attribute) (BedrockGenerator attribute) (Blank attribute) (BlankVision attribute) (CohereGenerator attribute) (Generator attribute) (GgmlGenerator attribute) (GroqChat attribute) (InferenceAPI attribute) (LangChainLLMGenerator attribute) (LangChainServeLLMGenerator attribute) (Lipsum attribute) (LiteLLMGenerator attribute) (LLMGenerator attribute) (MistralGenerator attribute) (Model attribute) (NeMoGuardrails attribute) (NeMoGuardrailsServer attribute) (Nones attribute) (NvcfChat attribute) (NVOpenAIChat attribute) (OllamaGenerator attribute) (OpenAICompatible attribute) (OpenAIGenerator attribute) (Pipeline attribute) (RasaRestGenerator attribute) (ReasoningLipsum attribute) (Repeat attribute) (ReplicateGenerator attribute) (RestGenerator attribute) (Single attribute), [1] (WatsonXGenerator attribute) generator_orig_tokens (Repeat attribute) GeneratorBackoffTrigger get_evaluations() (Report method) get_http_lib_agents() (in module garak._config) get_langprovider() (in module garak.langservice) get_z_rating() (Evaluator method) get_z_score() (Calibration method) getInstance() (PluginProvider static method) GetKey (class in garak.probes.apikey) GgmlGenerator (class in garak.generators.ggml) Glitch (class in garak.detectors.goodside) (class in garak.probes.glitch) glitch_tokens (GlitchFull attribute) GlitchFull (class in garak.probes.glitch) goal (Ablation_Dan_11_0 attribute) (AgentBreaker attribute) (AnsiEscaped attribute) (AnsiRaw attribute) (AnsiRawTokenizerHF attribute) (Anthropomorphisation attribute) (AntiDAN attribute) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (BadCharacters attribute) (BEAST attribute) (Blank attribute) (Bullying attribute) (Bypass attribute) (BypassLeet attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteKey attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlursFull attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWildFull attribute) (Dart attribute) (Davidjl attribute) (Deadnaming attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (DRA attribute) (DUDE attribute) (EICAR attribute) (EncodingMixin attribute) (Evasion attribute) (FactSnippetMixin attribute) (FalseAssertion attribute) (FigStepFull attribute) (FITD attribute) (FunctionMasking attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (GetKey attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivityFull attribute) (GTphish attribute) (GTUBE attribute) (HF_Files attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HomoglyphObfuscation attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (InformationHazard attribute) (JavaScript attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (MaliciousUses attribute) (MarkdownImageExfil attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (MisinformationHarms attribute) (MLM attribute) (PackageHallucinationProbe attribute) (PAIR attribute) (PastTenseFull attribute) (Payload attribute) (Perl attribute) (PIILeakProbeMixin attribute) (PIILeakQuadruplet attribute) (PIILeakTriplet attribute) (PIILeakTwin attribute) (PIILeakUnstructured attribute) (PlaygroundMarkdownExfil attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute) (Puppetry attribute) (Python attribute) (QuackMedicine attribute) (RakuLand attribute) (Repeat attribute) (RepeatedToken attribute), [1] (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (Ruby attribute) (Rust attribute) (SenatorsFull attribute) (SexualContent attribute) (Sexualisation attribute) (Slurs attribute) (SlurUsage attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (STAN attribute) (StringAssemblyDataExfil attribute) (SubFunctions attribute) (Substances attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (ThreatenJSON attribute) (TopLevel attribute) (Tox attribute) (TranslationMixin attribute) (WhoIsRiley attribute) (Win10 attribute) (Win11 attribute) (WordnetBlockedWords attribute) GOATAttack (class in garak.probes.goat) GraphConnectivity (class in garak.probes.snowball) GraphConnectivityFull (class in garak.probes.snowball) GroqChat (class in garak.generators.groq) GTphish (class in garak.detectors.knownbadsignatures) (class in garak.probes.av_spam_scanning) GTUBE (class in garak.detectors.knownbadsignatures) (class in garak.probes.av_spam_scanning) GuardianCloze (class in garak.probes.leakreplay) GuardianClozeFull (class in garak.probes.leakreplay) GuardianComplete (class in garak.probes.leakreplay) GuardianCompleteFull (class in garak.probes.leakreplay)

H

Harness (class in garak.harnesses.base) HF_Files (class in garak.probes.fileformats) HFDetector (class in garak.detectors.base) HFInternalServerError HFLoadingException HFRateLimitException HijackHateHumans (class in garak.probes.promptinject) HijackHateHumansFull (class in garak.probes.promptinject) HijackKillHumans (class in garak.probes.promptinject) HijackKillHumansFull (class in garak.probes.promptinject) HijackLongPrompt (class in garak.probes.promptinject) HijackLongPromptFull (class in garak.probes.promptinject) hint() (in module garak.command) hit_desc (Anthro attribute) (AntiDAN attribute) (AnyCode attribute) (ApiKey attribute) (Attack_On_Author attribute) (Attack_On_Commenter attribute) (AttackRogueString attribute) (Continuation attribute) (DAN attribute) (DANJailbreak attribute) (Dart attribute) (DecodeApprox attribute) (DecodeMatch attribute) (Detector attribute) (DevMode attribute) (DontStartNo attribute) (DontStartYes attribute) (Down attribute) (DUDE attribute) (EICAR attribute) (Escaped attribute) (FigStep attribute) (FileIsExecutable attribute) (FileIsPickled attribute) (Flirtation attribute) (Glitch attribute) (GTphish attribute) (GTUBE attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (Insult attribute) (Insult_Experimental attribute) (JavaScriptNpm attribute) (JinjaTemplateInjectionDetector attribute) (LDNOOBW attribute) (Likely_To_Reject attribute) (MarkdownExfilBasic attribute) (MarkdownExfilContent attribute) (MarkdownExfilExtendedImage attribute) (MarkdownExfilExtendedNonImage attribute) (MarkdownLink attribute) (Martin_haDistilbert attribute) (MitigationBypass attribute) (ModernBERTRefusal attribute) (MustContradictNLI attribute) (MustRefuteClaimKW attribute) (MustRefuteClaimModel attribute) (Obscene attribute) (OfcomOffensiveDiscriminatory attribute) (OfcomOffensiveMentalHealth attribute) (OfcomOffensiveRaceEthnic attribute) (OfcomOffensiveSexGender attribute) (OfcomOffensiveSexual attribute) (Perl attribute) (PlainJSON attribute) (PlaygroundMarkdownExfil attribute) (PossiblePickleName attribute) (Prefixes attribute) (Profanity attribute) (Profanity_Experimental attribute) (PromptExtraction attribute) (PromptExtractionVerbatim attribute) (PythonCodeExecution attribute) (PythonPypi attribute) (QuackMedicine attribute) (RakuLand attribute) (Raw attribute) (Refusal attribute) (RefusalOnlyAdversarial attribute) (RepeatDiverges attribute) (RepeatedToken attribute) (RileyIsnt attribute) (RubyGems attribute) (RustCrates attribute) (S_nlpDetox attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (Sexually_Explicit attribute) (SlursReclaimedSlurs attribute) (Spam attribute) (SQLiEcho attribute) (SQLiSuccess attribute) (STAN attribute) (StartsWith attribute) (SurgeProfanityAnimalReferences attribute) (SurgeProfanityBodilyFluids attribute) (SurgeProfanityMentalDisability attribute) (SurgeProfanityPhysicalAttributes attribute) (SurgeProfanityPhysicalDisability attribute) (SurgeProfanityPolitical attribute) (SurgeProfanityRacialEthnic attribute) (SurgeProfanityReligious attribute) (SurgeProfanitySexual attribute) (SurgeProfanitySexualOrientationGender attribute) (Threat attribute) (Threat_Experimental attribute) (ToxicCommentModel attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) (Up attribute) (Win5x5 attribute) (XSS attribute)

hitlogfile (TransientConfig attribute) HomoglyphObfuscation (class in garak.probes.smuggling) HumanChatbox (class in garak.probes.donotanswer) HypotheticalResponse (class in garak.probes.smuggling)

I

Identity_Attack (class in garak.detectors.perspective) Identity_Attack_Experimental (class in garak.detectors.perspective) Incoherent (class in garak.detectors.perspective) InferenceAPI (class in garak.generators.huggingface) InferenceEndpoint (class in garak.generators.huggingface) (class in garak.generators.replicate) Inflammatory (class in garak.detectors.perspective) INFORMATIONAL (Tier attribute) InformationHazard (class in garak.probes.donotanswer) InjectAscii85 (class in garak.probes.encoding) InjectAtbash (class in garak.probes.encoding) InjectBase16 (class in garak.probes.encoding) InjectBase2048 (class in garak.probes.encoding) InjectBase32 (class in garak.probes.encoding) InjectBase64 (class in garak.probes.encoding) InjectBraille (class in garak.probes.encoding) InjectEcoji (class in garak.probes.encoding) InjectHex (class in garak.probes.encoding) injection_instructions (FactSnippetMixin attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (LatentWhoisSnippetFull attribute) (TranslationMixin attribute)

injection_sep_pairs (LatentInjectionFactSnippetEiffelFull attribute) (LatentInjectionFactSnippetLegalFull attribute) (LatentInjectionMixin attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentWhois attribute) (TranslationMixin attribute) InjectLeet (class in garak.probes.encoding) InjectMime (class in garak.probes.encoding) InjectMorse (class in garak.probes.encoding) InjectNato (class in garak.probes.encoding) InjectQP (class in garak.probes.encoding) InjectROT13 (class in garak.probes.encoding) InjectSneakyBits (class in garak.probes.encoding) InjectUnicodeTagChars (class in garak.probes.encoding) InjectUnicodeVariantSelectors (class in garak.probes.encoding) InjectUU (class in garak.probes.encoding) InjectZalgo (class in garak.probes.encoding) instance() (PluginCache method) Insult (class in garak.detectors.perspective) Insult_Experimental (class in garak.detectors.perspective) interactive_mode() (in module garak.interactive) IterativeProbe (class in garak.probes.base)

J

Jailbreak (class in garak.detectors.judge) JailbreakOnlyAdversarial (class in garak.detectors.judge) JavaScript (class in garak.probes.packagehallucination)

JavaScriptNpm (class in garak.detectors.packagehallucination) JinjaTemplateInjectionDetector (class in garak.detectors.exploitation) JinjaTemplatePythonInjection (class in garak.probes.exploitation)

L

lang (Ablation_Dan_11_0 attribute) (AgentBreaker attribute) (AnsiEscaped attribute) (AnsiRawTokenizerHF attribute) (Anthropomorphisation attribute) (AntiDAN attribute) (Attempt property) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (BadCharacters attribute) (BEAST attribute) (Blank attribute) (Buff attribute) (Bullying attribute) (Bypass attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteKey attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlursFull attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWildFull attribute) (Davidjl attribute) (Deadnaming attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (DRA attribute) (DUDE attribute) (EICAR attribute) (EncodingMixin attribute) (Evasion attribute) (FactSnippetMixin attribute) (FalseAssertion attribute) (Fast attribute) (FigStepFull attribute) (FITD attribute) (FunctionMasking attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (GetKey attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivityFull attribute) (GTphish attribute) (GTUBE attribute) (HF_Files attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HomoglyphObfuscation attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (InformationHazard attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (MaliciousUses attribute) (MarkdownImageExfil attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (Message attribute) (MisinformationHarms attribute) (MLM attribute) (PackageHallucinationProbe attribute) (PAIR attribute) (PastTenseFull attribute) (Payload attribute) (PegasusT5 attribute) (PIILeakProbeMixin attribute) (PlaygroundMarkdownExfil attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute) (Puppetry attribute) (QuackMedicine attribute) (Repeat attribute) (RepeatedToken attribute) (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (SenatorsFull attribute) (SexualContent attribute) (Sexualisation attribute) (Slurs attribute) (SlurUsage attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (STAN attribute) (StringAssemblyDataExfil attribute) (SubFunctions attribute) (Substances attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (ThreatenJSON attribute) (TopLevel attribute) (Tox attribute) (TranslationMixin attribute) (WhoIsRiley attribute) (Win10 attribute) (WordnetBlockedWords attribute) lang_spec (AgentBreakerResult attribute) (Anthro attribute) (AntiDAN attribute) (AnyCode attribute) (AnyOutput attribute) (ApiKey attribute) (Attack_On_Author attribute) (Attack_On_Commenter attribute) (AttackRogueString attribute) (Continuation attribute) (DAN attribute) (DANJailbreak attribute) (DecodeApprox attribute) (DecodeMatch attribute) (Detector attribute) (DevMode attribute) (DontStartNo attribute) (DontStartYes attribute) (Down attribute) (DUDE attribute) (EICAR attribute) (Escaped attribute) (Fail attribute) (FigStep attribute) (FileIsExecutable attribute) (FileIsPickled attribute) (Flirtation attribute) (Glitch attribute) (GTphish attribute) (GTUBE attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (Insult attribute) (Insult_Experimental attribute) (JinjaTemplateInjectionDetector attribute) (LDNOOBW attribute) (Likely_To_Reject attribute) (MarkdownExfilBasic attribute) (MarkdownExfilContent attribute) (MarkdownExfilExtendedImage attribute) (MarkdownExfilExtendedNonImage attribute) (MarkdownLink attribute) (Martin_haDistilbert attribute) (MitigationBypass attribute) (ModelAsJudge attribute) (ModernBERTRefusal attribute) (MustContradictNLI attribute) (MustRefuteClaimKW attribute) (MustRefuteClaimModel attribute) (Obscene attribute) (OfcomOffensiveDiscriminatory attribute) (OfcomOffensiveGeneral attribute) (OfcomOffensiveMentalHealth attribute) (OfcomOffensiveRaceEthnic attribute) (OfcomOffensiveSexGender attribute) (OfcomOffensiveSexual attribute) (PackageHallucinationDetector attribute) (Pass attribute) (Passthru attribute) (PIILeak attribute) (PIILeakExact attribute) (PlainJSON attribute) (PossiblePickleName attribute) (Prefixes attribute) (Profanity attribute) (Profanity_Experimental attribute) (PromptExtraction attribute) (PythonCodeExecution attribute) (QuackMedicine attribute) (Random attribute) (Raw attribute) (RepeatDiverges attribute) (RepeatedToken attribute) (RileyIsnt attribute) (S_nlpDetox attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (Sexually_Explicit attribute) (SlursReclaimedSlurs attribute) (Spam attribute) (SQLiEcho attribute) (SQLiSuccess attribute) (STAN attribute) (StartsWith attribute) (SurgeProfanityAnimalReferences attribute) (SurgeProfanityBodilyFluids attribute) (SurgeProfanityMentalDisability attribute) (SurgeProfanityPhysicalAttributes attribute) (SurgeProfanityPhysicalDisability attribute) (SurgeProfanityPolitical attribute) (SurgeProfanityRacialEthnic attribute) (SurgeProfanityReligious attribute) (SurgeProfanitySexual attribute) (SurgeProfanitySexualOrientationGender attribute) (Threat attribute) (Threat_Experimental attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) (Up attribute) (Win5x5 attribute) (XSS attribute)

LangChainLLMGenerator (class in garak.generators.langchain) LangChainServeLLMGenerator (class in garak.generators.langchain_serve) language_name (Dart attribute), [1] (JavaScript attribute) (JavaScriptNpm attribute) (PackageHallucinationDetector property) (PackageHallucinationProbe property) (Perl attribute), [1] (Python attribute) (PythonPypi attribute) (RakuLand attribute), [1] (Ruby attribute) (RubyGems attribute) (Rust attribute) (RustCrates attribute) last_message() (Conversation method) LatentInjectionFactSnippetEiffel (class in garak.probes.latentinjection) LatentInjectionFactSnippetEiffelFull (class in garak.probes.latentinjection) LatentInjectionFactSnippetLegal (class in garak.probes.latentinjection) LatentInjectionFactSnippetLegalFull (class in garak.probes.latentinjection) LatentInjectionMixin (class in garak.probes.latentinjection) LatentInjectionReport (class in garak.probes.latentinjection) LatentInjectionReportFull (class in garak.probes.latentinjection) LatentInjectionResume (class in garak.probes.latentinjection) LatentInjectionResumeFull (class in garak.probes.latentinjection) LatentInjectionTranslationEnFr (class in garak.probes.latentinjection) LatentInjectionTranslationEnFrFull (class in garak.probes.latentinjection) LatentInjectionTranslationEnZh (class in garak.probes.latentinjection) LatentInjectionTranslationEnZhFull (class in garak.probes.latentinjection) LatentJailbreak (class in garak.probes.latentinjection) LatentJailbreakFull (class in garak.probes.latentinjection) LatentWhois (class in garak.probes.latentinjection) LatentWhoisSnippet (class in garak.probes.latentinjection) LatentWhoisSnippetFull (class in garak.probes.latentinjection) LDNOOBW (class in garak.detectors.unsafe_content) leet_bytes() (in module garak.probes.encoding) Likely_To_Reject (class in garak.detectors.perspective) Lipsum (class in garak.generators.test) list_config() (in module garak.command) LiteLLMGenerator (class in garak.generators.litellm) LiteratureCloze (class in garak.probes.leakreplay) LiteratureClozeFull (class in garak.probes.leakreplay) LiteratureComplete (class in garak.probes.leakreplay) LiteratureCompleteFull (class in garak.probes.leakreplay) LLaVA (class in garak.generators.huggingface) LLMGenerator (class in garak.generators.llm) load() (Director method) (in module garak.langservice) (in module garak.payloads) (Report method) load_base_config() (in module garak._config) load_config() (in module garak._config) load_data() (DiscriminationExclusionToxicityHatefulOffensive method) (HumanChatbox method) (InformationHazard method) (MaliciousUses method) (MisinformationHarms method) (RTPFlirtation method) (RTPIdentity_Attack method) (RTPInsult method) (RTPProfanity method) (RTPSevere_Toxicity method) (RTPSexually_Explicit method) (RTPThreat method) load_data_all_attacks() (HijackHateHumans method) (HijackHateHumansFull method) (HijackKillHumans method) (HijackKillHumansFull method) (HijackLongPrompt method) (HijackLongPromptFull method) (in module garak.probes.promptinject) load_local_data() (in module garak.probes.donotanswer) (in module garak.probes.realtoxicityprompts) load_plugin() (in module garak._plugins) local_constructor() (in module garak.probes.donotanswer) (in module garak.probes.realtoxicityprompts) log_filename (TransientConfig attribute) Lowercase (class in garak.buffs.lowercase) LRLBuff (class in garak.buffs.low_resource_languages)

M

main() (in module garak.cli) MaliciousUses (class in garak.probes.donotanswer) markdown_uri_templates (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) MarkdownExfilBasic (class in garak.detectors.web_injection) MarkdownExfilContent (class in garak.detectors.web_injection) MarkdownExfilExtendedImage (class in garak.detectors.web_injection) MarkdownExfilExtendedNonImage (class in garak.detectors.web_injection) MarkdownExfilMixin (class in garak.probes.web_injection) MarkdownImageExfil (class in garak.probes.web_injection) MarkdownLink (class in garak.detectors.dan) MarkdownURIImageExfilExtended (class in garak.probes.web_injection) MarkdownURINonImageExfilExtended (class in garak.probes.web_injection) MarkdownXSS (class in garak.probes.web_injection) Martin_haDistilbert (class in garak.detectors.unsafe_content) matchtype (Anthro attribute) (AntiDAN attribute) (DAN attribute) (DANJailbreak attribute) (DevMode attribute) (DUDE attribute) (EICAR attribute) (FigStep attribute) (GTphish attribute) (GTUBE attribute) (MarkdownLink attribute) (QuackMedicine attribute) (RileyIsnt attribute) (STAN attribute) max_prompts (GlitchFull attribute) Message (class in garak.attempt) MisinformationHarms (class in garak.probes.donotanswer) MistralGenerator (class in garak.generators.mistral) MitigationBypass (class in garak.detectors.mitigation) MLM (class in garak.probes.sata) modality (AudioAchillesHeel attribute) (BlankVision attribute) (Detector attribute) (FigStepFull attribute) (Generator attribute) (HF_Files attribute) (LLaVA attribute) (NVMultimodal attribute) (Probe attribute) (Vision attribute) Model (class in garak.generators.huggingface) MODEL_NAME_ENV_VAR (AzureOpenAIGenerator attribute) ModelAsJudge (class in garak.detectors.judge) ModernBERTRefusal (class in garak.detectors.mitigation) module garak._config garak._plugins garak.analyze.calibration garak.analyze.report_digest garak.attempt garak.buffs.base garak.buffs.encoding garak.buffs.low_resource_languages garak.buffs.lowercase garak.buffs.paraphrase garak.cli garak.command garak.detectors.agent_breaker garak.detectors.always garak.detectors.ansiescape garak.detectors.any garak.detectors.apikey garak.detectors.base garak.detectors.continuation garak.detectors.dan garak.detectors.divergence garak.detectors.encoding garak.detectors.exploitation garak.detectors.fileformats garak.detectors.goodside garak.detectors.judge garak.detectors.knownbadsignatures garak.detectors.leakreplay garak.detectors.lmrc garak.detectors.malwaregen garak.detectors.misleading garak.detectors.mitigation garak.detectors.packagehallucination garak.detectors.perspective garak.detectors.productkey garak.detectors.promptinject garak.detectors.propile garak.detectors.shields garak.detectors.snowball garak.detectors.sysprompt_extraction garak.detectors.unsafe_content garak.detectors.visual_jailbreak garak.detectors.web_injection garak.evaluators.base garak.exception garak.generators.anthropic garak.generators.azure garak.generators.base garak.generators.bedrock garak.generators.cohere garak.generators.function garak.generators.ggml garak.generators.groq garak.generators.guardrails garak.generators.huggingface garak.generators.langchain garak.generators.langchain_serve garak.generators.litellm garak.generators.llm garak.generators.mistral garak.generators.nim garak.generators.nvcf garak.generators.ollama garak.generators.openai garak.generators.rasa garak.generators.replicate garak.generators.rest garak.generators.test garak.generators.watsonx garak.generators.websocket garak.harnesses.base garak.harnesses.probewise garak.harnesses.pxd garak.interactive garak.langservice garak.payloads garak.probes._tier garak.probes.agent_breaker garak.probes.ansiescape garak.probes.apikey garak.probes.atkgen garak.probes.audio garak.probes.av_spam_scanning garak.probes.badchars garak.probes.base garak.probes.continuation garak.probes.dan garak.probes.divergence garak.probes.doctor garak.probes.donotanswer garak.probes.dra garak.probes.encoding garak.probes.exploitation garak.probes.fileformats garak.probes.fitd garak.probes.glitch garak.probes.goat garak.probes.goodside garak.probes.grandma garak.probes.latentinjection garak.probes.leakreplay garak.probes.lmrc garak.probes.malwaregen garak.probes.misleading garak.probes.packagehallucination garak.probes.phrasing garak.probes.promptinject garak.probes.propile garak.probes.realtoxicityprompts garak.probes.sata garak.probes.smuggling garak.probes.snowball garak.probes.suffix garak.probes.sysprompt_extraction garak.probes.tap garak.probes.test garak.probes.topic garak.probes.visual_jailbreak garak.probes.web_injection garak.report

morse() (in module garak.probes.encoding) Multiple (class in garak.generators.function) MustContradictNLI (class in garak.detectors.misleading) MustRefuteClaimKW (class in garak.detectors.misleading) MustRefuteClaimModel (class in garak.detectors.misleading) mutate_domain() (ColabAIDataLeakage static method)

N

name (Blank attribute) (BlankVision attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (Lipsum attribute) (Nones attribute) (ReasoningLipsum attribute) (Repeat attribute) (RepeatedToken attribute) (Single attribute) nato() (InjectNato method) NeMoGuardrails (class in garak.generators.guardrails) NeMoGuardrailsServer (class in garak.generators.guardrails) nested_dict() (in module garak._config)

never_queue_forms (TreeSearchProbe attribute) never_queue_nodes (TreeSearchProbe attribute) new_max_tokens (Repeat attribute) non_full_suffix (NonFullMixin attribute) non_injection_text (LatentWhoisSnippetFull attribute) Nones (class in garak.generators.test) NonFullMixin (class in garak.probes.latentinjection) (class in garak.probes.leakreplay) notes (Conversation attribute) (Message attribute) NvcfChat (class in garak.generators.nvcf) NvcfCompletion (class in garak.generators.nvcf) NVMultimodal (class in garak.generators.nim) NVOpenAIChat (class in garak.generators.nim) NVOpenAICompletion (class in garak.generators.nim) NYTCloze (class in garak.probes.leakreplay) NYTClozeFull (class in garak.probes.leakreplay) NYTComplete (class in garak.probes.leakreplay) NYTCompleteFull (class in garak.probes.leakreplay)

O

Obscene (class in garak.detectors.perspective) OF_CONCERN (Tier attribute) OfcomOffensiveDiscriminatory (class in garak.detectors.unsafe_content) OfcomOffensiveGeneral (class in garak.detectors.unsafe_content) OfcomOffensiveMentalHealth (class in garak.detectors.unsafe_content) OfcomOffensiveRaceEthnic (class in garak.detectors.unsafe_content) OfcomOffensiveSexGender (class in garak.detectors.unsafe_content) OfcomOffensiveSexual (class in garak.detectors.unsafe_content)

OllamaGenerator (class in garak.generators.ollama) OllamaGeneratorChat (class in garak.generators.ollama) OpenAICompatible (class in garak.generators.openai) OpenAIGenerator (class in garak.generators.openai) OpenAIReasoningGenerator (class in garak.generators.openai) outputs (Attempt property) outputs_for() (Attempt method) override_maxlen (Repeat attribute)

P

package_dir (TransientConfig attribute) PackageHallucinationDetector (class in garak.detectors.packagehallucination) PackageHallucinationProbe (class in garak.probes.packagehallucination) packages (PackageHallucinationDetector attribute) PAIR (class in garak.probes.tap) paragraphs (LatentInjectionFactSnippetEiffelFull attribute) (LatentInjectionFactSnippetLegalFull attribute) (LatentWhoisSnippetFull attribute) parallel_capable (Generator attribute) (LLaVA attribute) (LLMGenerator attribute) (OllamaGenerator attribute) (Pipeline attribute) parallelisable_attempts (AgentBreaker attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (Probe attribute) (Tox attribute) parse_cli_plugin_config() (in module garak.cli) parse_plugin_spec() (in module garak._config) partial_keys (CompleteKey attribute) partial_match (PIILeak attribute) partial_threshold (PIILeak attribute) Pass (class in garak.detectors.always) pass_desc (Anthro attribute) (AntiDAN attribute) (AnyCode attribute) (ApiKey attribute) (Attack_On_Author attribute) (Attack_On_Commenter attribute) (AttackRogueString attribute) (Continuation attribute) (DAN attribute) (DANJailbreak attribute) (Dart attribute) (DecodeApprox attribute) (DecodeMatch attribute) (Detector attribute) (DevMode attribute) (DontStartNo attribute) (DontStartYes attribute) (Down attribute) (DUDE attribute) (EICAR attribute) (Escaped attribute) (FigStep attribute) (FileIsExecutable attribute) (FileIsPickled attribute) (Flirtation attribute) (Glitch attribute) (GTphish attribute) (GTUBE attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (Insult attribute) (Insult_Experimental attribute) (JavaScriptNpm attribute) (JinjaTemplateInjectionDetector attribute) (LDNOOBW attribute) (Likely_To_Reject attribute) (MarkdownExfilBasic attribute) (MarkdownExfilContent attribute) (MarkdownExfilExtendedImage attribute) (MarkdownExfilExtendedNonImage attribute) (MarkdownLink attribute) (Martin_haDistilbert attribute) (MitigationBypass attribute) (ModernBERTRefusal attribute) (MustContradictNLI attribute) (MustRefuteClaimKW attribute) (MustRefuteClaimModel attribute) (Obscene attribute) (OfcomOffensiveDiscriminatory attribute) (OfcomOffensiveMentalHealth attribute) (OfcomOffensiveRaceEthnic attribute) (OfcomOffensiveSexGender attribute) (OfcomOffensiveSexual attribute) (Perl attribute) (PlainJSON attribute) (PlaygroundMarkdownExfil attribute) (PossiblePickleName attribute) (Prefixes attribute) (Profanity attribute) (Profanity_Experimental attribute) (PromptExtraction attribute) (PromptExtractionVerbatim attribute) (PythonCodeExecution attribute) (PythonPypi attribute) (QuackMedicine attribute) (RakuLand attribute) (Raw attribute) (Refusal attribute) (RefusalOnlyAdversarial attribute) (RepeatDiverges attribute) (RepeatedToken attribute) (RileyIsnt attribute) (RubyGems attribute) (RustCrates attribute) (S_nlpDetox attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (Sexually_Explicit attribute) (SlursReclaimedSlurs attribute) (Spam attribute) (SQLiEcho attribute) (SQLiSuccess attribute) (STAN attribute) (StartsWith attribute) (SurgeProfanityAnimalReferences attribute) (SurgeProfanityBodilyFluids attribute) (SurgeProfanityMentalDisability attribute) (SurgeProfanityPhysicalAttributes attribute) (SurgeProfanityPhysicalDisability attribute) (SurgeProfanityPolitical attribute) (SurgeProfanityRacialEthnic attribute) (SurgeProfanityReligious attribute) (SurgeProfanitySexual attribute) (SurgeProfanitySexualOrientationGender attribute) (Threat attribute) (Threat_Experimental attribute) (ToxicCommentModel attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) (Up attribute) (Win5x5 attribute) (XSS attribute) Passthru (class in garak.detectors.always) PastTense (class in garak.probes.phrasing) PastTenseFull (class in garak.probes.phrasing) Payload (class in garak.probes.malwaregen) payload (Win10 attribute) (Win11 attribute) payload_list (Director attribute) payload_template (JinjaTemplatePythonInjection attribute) payload_triggers (LatentInjectionFactSnippetEiffelFull attribute) (LatentInjectionFactSnippetLegalFull attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (LatentWhoisSnippetFull attribute) (Tag attribute) (TranslationMixin attribute) PayloadFailure PayloadGroup (class in garak.payloads) payloads (FactSnippetMixin attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (Tag attribute) (TranslationMixin attribute) (XSS attribute) PegasusT5 (class in garak.buffs.paraphrase) Perl (class in garak.detectors.packagehallucination) (class in garak.probes.packagehallucination) PID_ENV_VAR (WatsonXGenerator attribute) PIILeak (class in garak.detectors.propile) PIILeakExact (class in garak.detectors.propile) PIILeakProbeMixin (class in garak.probes.propile) PIILeakQuadruplet (class in garak.probes.propile) PIILeakTriplet (class in garak.probes.propile) PIILeakTwin (class in garak.probes.propile) PIILeakUnstructured (class in garak.probes.propile) Pipeline (class in garak.generators.huggingface) PlainJSON (class in garak.detectors.goodside) PlaygroundMarkdownExfil (class in garak.detectors.web_injection) (class in garak.probes.web_injection) plugin_docstring_to_description() (in module garak.analyze.report_digest) plugin_info() (in module garak._plugins) (in module garak.command) (PluginCache method) PluginCache (class in garak._plugins) PluginConfigurationError PluginEncoder (class in garak._plugins) PluginProvider (class in garak._plugins) PossiblePickleName (class in garak.detectors.fileformats) post_buff_hook (Probe attribute) postcmd() (GarakTerminal method) PotterCloze (class in garak.probes.leakreplay) PotterClozeFull (class in garak.probes.leakreplay) PotterComplete (class in garak.probes.leakreplay) PotterCompleteFull (class in garak.probes.leakreplay) Prefixes (class in garak.detectors.mitigation) primary_detector (Ablation_Dan_11_0 attribute) (AgentBreaker attribute) (AnsiEscaped attribute) (AnsiRaw attribute) (AnsiRawTokenizerHF attribute) (Anthropomorphisation attribute) (AntiDAN attribute) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (BadCharacters attribute) (BEAST attribute) (Blank attribute) (Bullying attribute) (Bypass attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteKey attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlursFull attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWildFull attribute) (Dart attribute) (Davidjl attribute) (Deadnaming attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (DRA attribute) (DUDE attribute) (EICAR attribute) (EncodingMixin attribute) (Evasion attribute) (FalseAssertion attribute) (FigStepFull attribute) (FITD attribute) (FunctionMasking attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (GetKey attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivityFull attribute) (GTphish attribute) (GTUBE attribute) (HF_Files attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HomoglyphObfuscation attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (InformationHazard attribute) (JavaScript attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionMixin attribute) (LatentJailbreakFull attribute) (MaliciousUses attribute) (MarkdownImageExfil attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (MisinformationHarms attribute) (MLM attribute) (PAIR attribute) (PastTenseFull attribute) (Payload attribute) (Perl attribute) (PIILeakProbeMixin attribute) (PlaygroundMarkdownExfil attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute) (Puppetry attribute) (Python attribute) (QuackMedicine attribute) (RakuLand attribute) (Repeat attribute) (RepeatedToken attribute) (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (Ruby attribute) (Rust attribute) (SenatorsFull attribute) (SexualContent attribute) (Sexualisation attribute) (Slurs attribute) (SlurUsage attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (STAN attribute) (StringAssemblyDataExfil attribute) (SubFunctions attribute) (Substances attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (ThreatenJSON attribute) (TopLevel attribute) (Tox attribute) (WhoIsRiley attribute) (Win10 attribute) (WordnetBlockedWords attribute)

Primes (class in garak.probes.snowball) PrimesFull (class in garak.probes.snowball) print_buffs() (in module garak.command) print_detectors() (in module garak.command) print_generators() (in module garak.command) print_plugins() (in module garak.command) (in module garak.interactive) print_probes() (in module garak.command) print_results_narrow() (Evaluator method) print_results_wide() (Evaluator method) Probe (class in garak.probes.base) probe() (Ablation_Dan_11_0 method) (AnsiRawTokenizerHF method) (AntiDAN method) (AudioAchillesHeel method) (AutoDAN method) (AutoDANCached method) (BEAST method) (ChatGPT_Developer_Mode_RANTI method) (ChatGPT_Developer_Mode_v2 method) (ChatGPT_Image_Markdown method) (Dan_10_0 method) (Dan_11_0 method) (Dan_6_0 method) (Dan_6_2 method) (Dan_7_0 method) (Dan_8_0 method) (Dan_9_0 method) (DAN_Jailbreak method) (DanInTheWild method) (DanInTheWildFull method) (DUDE method) (FigStep method) (FigStepFull method) (GCG method) (HF_Files method) (IterativeProbe method) (PAIR method) (PIILeakProbeMixin method) (Probe method) (STAN method) (TAP method) (Tox method) (TreeSearchProbe method) probewise_run() (in module garak.command) ProbewiseHarness (class in garak.harnesses.probewise) prodkey_5x5_regex (Win5x5 attribute) Profanity (class in garak.detectors.perspective) (class in garak.probes.lmrc) Profanity_Experimental (class in garak.detectors.perspective) prompt (Attempt property) (FunctionMasking attribute) prompt1 (GCGCached attribute) prompt2 (GCGCached attribute) prompt_file (Ablation_Dan_11_0 attribute) (AntiDAN attribute) (AutoDAN attribute) (AutoDANCached attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWild attribute) (DanInTheWildFull attribute) (DUDE attribute) (STAN attribute) prompt_for() (Attempt method) PromptExtraction (class in garak.detectors.sysprompt_extraction) PromptExtractionVerbatim (class in garak.detectors.sysprompt_extraction) prompts (Anthropomorphisation attribute) (BEAST attribute) (Blank attribute) (Bullying attribute) (Davidjl attribute) (Deadnaming attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (FigStepFull attribute) (FunctionMasking attribute) (GCG attribute) (GCGCached attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (InformationHazard attribute) (MaliciousUses attribute) (MisinformationHarms attribute) (Profanity attribute) (QuackMedicine attribute) (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (SexualContent attribute) (Sexualisation attribute) (SlurUsage attribute) (Test attribute) (ThreatenJSON attribute) (WhoIsRiley attribute) Puppetry (class in garak.probes.doctor) PxD (class in garak.harnesses.pxd) pxd_run() (in module garak.command) Python (class in garak.probes.packagehallucination) PythonCodeExecution (class in garak.detectors.exploitation) PythonPypi (class in garak.detectors.packagehallucination)

Q

QuackMedicine (class in garak.detectors.lmrc) (class in garak.probes.lmrc)

R

RakuLand (class in garak.detectors.packagehallucination) (class in garak.probes.packagehallucination) Random (class in garak.detectors.always) RasaRestGenerator (class in garak.generators.rasa) RateLimitHit Raw (class in garak.detectors.ansiescape) ReasoningLipsum (class in garak.generators.test) recommended_detector (Probe attribute) Refusal (class in garak.detectors.judge) RefusalOnlyAdversarial (class in garak.detectors.judge) regex (JinjaTemplateInjectionDetector attribute) (MarkdownExfilBasic attribute) (PlaygroundMarkdownExfil attribute) (SQLiEcho attribute) relationships (PIILeakUnstructured attribute) Repeat (class in garak.generators.test) (class in garak.probes.divergence) repeat_word_list (Repeat attribute) (RepeatExtended attribute) RepeatDiverges (class in garak.detectors.divergence) REPEATED_SAMPLES_REGEX (RepeatedToken attribute) RepeatedToken (class in garak.detectors.divergence) (class in garak.probes.divergence) RepeatExtended (class in garak.probes.divergence) ReplicateGenerator (class in garak.generators.replicate) Report (class in garak.report) report_filename (TransientConfig attribute) reportfile (TransientConfig attribute)

ReportIncompatibleError requests (InferenceAPI attribute) (InferenceEndpoint attribute) RestGenerator (class in garak.generators.rest) RileyIsnt (class in garak.detectors.goodside) rogue_string (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) role (Turn attribute) rot13() (in module garak.probes.encoding) RTPBlank (class in garak.probes.realtoxicityprompts) RTPFlirtation (class in garak.probes.realtoxicityprompts) RTPIdentity_Attack (class in garak.probes.realtoxicityprompts) RTPInsult (class in garak.probes.realtoxicityprompts) RTPProfanity (class in garak.probes.realtoxicityprompts) RTPSevere_Toxicity (class in garak.probes.realtoxicityprompts) RTPSexually_Explicit (class in garak.probes.realtoxicityprompts) RTPThreat (class in garak.probes.realtoxicityprompts) Ruby (class in garak.probes.packagehallucination) RubyGems (class in garak.detectors.packagehallucination) run() (Harness method) (ProbewiseHarness method) (PxD method) run_id (TransientConfig attribute) Rust (class in garak.probes.packagehallucination) RustCrates (class in garak.detectors.packagehallucination)

S

S_nlpDetox (class in garak.detectors.unsafe_content) safebench_image_catalog (FigStep attribute) (FigStepFull attribute) safebench_image_filenames (FigStepFull attribute) search() (Director method) (in module garak.payloads) Senators (class in garak.probes.snowball) SenatorsFull (class in garak.probes.snowball) set_all_http_lib_agents() (in module garak._config) set_http_lib_agents() (in module garak._config) set_up_attack_gen_llm() (FITD method) set_up_detector() (FITD method) set_up_refusal_detector() (FITD method) settings_ns_provider() (GarakTerminal method) Severe_Toxicity (class in garak.detectors.perspective) Severe_Toxicity_Experimental (class in garak.detectors.perspective) SexualContent (class in garak.probes.lmrc) Sexualisation (class in garak.probes.lmrc) Sexually_Explicit (class in garak.detectors.perspective) Single (class in garak.generators.function) (class in garak.generators.test) Slurs (class in garak.probes.grandma) SlursReclaimedSlurs (class in garak.detectors.unsafe_content) SlurUsage (class in garak.probes.lmrc) source_file (GuardianClozeFull attribute) (GuardianCompleteFull attribute) (LiteratureClozeFull attribute) (LiteratureCompleteFull attribute) (NYTClozeFull attribute) (NYTCompleteFull attribute) (PotterClozeFull attribute) (PotterCompleteFull attribute) source_resource_filename (ContinueSlursReclaimedSlurs attribute) (ContinueSlursReclaimedSlursFull attribute) Spam (class in garak.detectors.perspective) SQLiEcho (class in garak.detectors.exploitation) SQLInjectionEcho (class in garak.probes.exploitation) SQLInjectionSystem (class in garak.probes.exploitation) SQLiSuccess (class in garak.detectors.exploitation) STAN (class in garak.detectors.dan) (class in garak.probes.dan) start_logging() (in module garak.command) start_msg() (in module garak.langservice) start_run() (in module garak.command) StartsWith (class in garak.detectors.leakreplay) starttime (TransientConfig attribute) starttime_iso (TransientConfig attribute) storeInstance() (PluginProvider static method) StringAssemblyDataExfil (class in garak.probes.web_injection) StringDetector (class in garak.detectors.base) SubFunctions (class in garak.probes.malwaregen)

Substances (class in garak.probes.grandma) substrings (AntiDAN attribute) (DAN attribute) (DANJailbreak attribute) (DevMode attribute) (DUDE attribute) (MarkdownLink attribute) (SQLiSuccess attribute) (STAN attribute) suffix (GCGCached attribute) suffixes (GCGCached attribute) supported_generators (AnsiRawTokenizerHF attribute) (HF_Files attribute) supported_models (LLaVA attribute) supports_multiple_generations (AnthropicGenerator attribute) (BedrockGenerator attribute) (Blank attribute) (BlankVision attribute) (Generator attribute) (GroqChat attribute) (InferenceAPI attribute) (InferenceEndpoint attribute) (Lipsum attribute) (LiteLLMGenerator attribute) (MistralGenerator attribute) (Model attribute) (Multiple attribute) (NeMoGuardrails attribute) (NeMoGuardrailsServer attribute) (Nones attribute) (NvcfChat attribute) (NVOpenAIChat attribute) (OpenAICompatible attribute) (OpenAIGenerator attribute) (OpenAIReasoningGenerator attribute) (Pipeline attribute) (ReasoningLipsum attribute) (Repeat attribute) (ReplicateGenerator attribute) (Single attribute), [1] SurgeProfanityAnimalReferences (class in garak.detectors.unsafe_content) SurgeProfanityBodilyFluids (class in garak.detectors.unsafe_content) SurgeProfanityMentalDisability (class in garak.detectors.unsafe_content) SurgeProfanityPhysicalAttributes (class in garak.detectors.unsafe_content) SurgeProfanityPhysicalDisability (class in garak.detectors.unsafe_content) SurgeProfanityPolitical (class in garak.detectors.unsafe_content) SurgeProfanityRacialEthnic (class in garak.detectors.unsafe_content) SurgeProfanityReligious (class in garak.detectors.unsafe_content) SurgeProfanitySexual (class in garak.detectors.unsafe_content) SurgeProfanitySexualOrientationGender (class in garak.detectors.unsafe_content) SYMBOL_SET (Evaluator attribute) SystemPromptExtraction (class in garak.probes.sysprompt_extraction)

T

Tag (class in garak.probes.goodside) tags (Ablation_Dan_11_0 attribute) (AgentBreaker attribute) (AgentBreakerResult attribute) (AnsiEscaped attribute) (AnsiRawTokenizerHF attribute) (Anthro attribute) (Anthropomorphisation attribute) (AntiDAN attribute), [1] (AnyOutput attribute) (Attack_On_Author attribute) (Attack_On_Commenter attribute) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (BadCharacters attribute) (BEAST attribute) (Blank attribute) (Bullying attribute) (Bypass attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteKey attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlursFull attribute) (DAN attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWildFull attribute) (DANJailbreak attribute) (Davidjl attribute) (Deadnaming attribute) (Detector attribute) (DevMode attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (DontStartNo attribute) (DontStartYes attribute) (DRA attribute) (DUDE attribute), [1] (EICAR attribute), [1] (EncodingMixin attribute) (Evasion attribute) (FalseAssertion attribute) (FigStep attribute) (FigStepFull attribute) (FITD attribute) (Flirtation attribute) (FunctionMasking attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (GetKey attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivityFull attribute) (GTphish attribute), [1] (GTUBE attribute), [1] (HF_Files attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HomoglyphObfuscation attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (Identity_Attack attribute) (Identity_Attack_Experimental attribute) (Incoherent attribute) (Inflammatory attribute) (InformationHazard attribute) (Insult attribute) (Insult_Experimental attribute) (JinjaTemplateInjectionDetector attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionMixin attribute) (LDNOOBW attribute) (Likely_To_Reject attribute) (MaliciousUses attribute) (MarkdownExfilBasic attribute) (MarkdownExfilContent attribute) (MarkdownExfilExtendedImage attribute) (MarkdownExfilExtendedNonImage attribute) (MarkdownImageExfil attribute) (MarkdownLink attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (Martin_haDistilbert attribute) (MisinformationHarms attribute) (MitigationBypass attribute) (MLM attribute) (ModernBERTRefusal attribute) (Obscene attribute) (OfcomOffensiveDiscriminatory attribute) (OfcomOffensiveGeneral attribute) (OfcomOffensiveMentalHealth attribute) (OfcomOffensiveRaceEthnic attribute) (OfcomOffensiveSexGender attribute) (OfcomOffensiveSexual attribute) (PackageHallucinationProbe attribute) (PAIR attribute) (PastTenseFull attribute) (Payload attribute) (PIILeakProbeMixin attribute) (PlaygroundMarkdownExfil attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute), [1] (Profanity_Experimental attribute) (PromptExtraction attribute) (Puppetry attribute) (PythonCodeExecution attribute) (QuackMedicine attribute), [1] (Repeat attribute) (RepeatedToken attribute), [1] (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (S_nlpDetox attribute) (SenatorsFull attribute) (Severe_Toxicity attribute) (Severe_Toxicity_Experimental attribute) (SexualContent attribute) (Sexualisation attribute) (Sexually_Explicit attribute) (Slurs attribute) (SlursReclaimedSlurs attribute) (SlurUsage attribute) (Spam attribute) (SQLiEcho attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (SQLiSuccess attribute) (STAN attribute), [1] (StringAssemblyDataExfil attribute) (SubFunctions attribute) (Substances attribute) (SurgeProfanityAnimalReferences attribute) (SurgeProfanityBodilyFluids attribute) (SurgeProfanityMentalDisability attribute) (SurgeProfanityPhysicalAttributes attribute) (SurgeProfanityPhysicalDisability attribute) (SurgeProfanityPolitical attribute) (SurgeProfanityRacialEthnic attribute) (SurgeProfanityReligious attribute) (SurgeProfanitySexual attribute) (SurgeProfanitySexualOrientationGender attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (Threat attribute) (Threat_Experimental attribute) (ThreatenJSON attribute) (TopLevel attribute) (Tox attribute) (Toxicity attribute) (Toxicity_Experimental attribute) (Unsubstantial attribute) (WhoIsRiley attribute) (Win10 attribute) (WordnetBlockedWords attribute) (XSS attribute) TAP (class in garak.probes.tap) TAPCached (class in garak.probes.tap) target_lang_name_en (LatentInjectionTranslationEnFrFull attribute) (LatentInjectionTranslationEnZhFull attribute) (TranslationMixin attribute) TargetNameMissingError tasks (Payload attribute) tasks() (in module garak.langservice) TaskXSS (class in garak.probes.web_injection) template (ColabAIDataLeakage attribute) (JinjaTemplatePythonInjection attribute) (MarkdownImageExfil attribute) (MLM attribute) (PlaygroundMarkdownExfil attribute) (SQLInjectionEcho attribute) (StringAssemblyDataExfil attribute) (TaskXSS attribute)

templates (GlitchFull attribute) TEMPLATES (WordnetBlockedWords attribute) term (OfcomOffensiveDiscriminatory attribute) (OfcomOffensiveGeneral attribute) (OfcomOffensiveMentalHealth attribute) (OfcomOffensiveRaceEthnic attribute) (OfcomOffensiveSexGender attribute) (OfcomOffensiveSexual attribute) Test (class in garak.probes.test) test() (Evaluator method) (ThresholdEvaluator method) (ZeroToleranceEvaluator method) test_generation_string (Single attribute) test_regexes (MarkdownExfilExtendedImage attribute) (MarkdownExfilExtendedNonImage attribute) text (Message attribute) Threat (class in garak.detectors.perspective) Threat_Experimental (class in garak.detectors.perspective) ThreatenJSON (class in garak.probes.goodside) threshold (AnyCode attribute) (MarkdownExfilContent attribute) ThresholdEvaluator (class in garak.evaluators.base) tier (Ablation_Dan_11_0 attribute) (AgentBreaker attribute) (AnsiEscaped attribute) (AnsiRaw attribute) (AnsiRawTokenizerHF attribute) (Anthropomorphisation attribute) (AntiDAN attribute) (AudioAchillesHeel attribute) (AutoDAN attribute) (AutoDANCached attribute) (BadCharacters attribute) (BEAST attribute) (Blank attribute) (Bullying attribute) (Bypass attribute) (BypassLeet attribute) (ChatGPT_Developer_Mode_RANTI attribute) (ChatGPT_Developer_Mode_v2 attribute) (ChatGPT_Image_Markdown attribute) Tier (class in garak.probes._tier) tier (ClozeProbeMixin attribute) (ColabAIDataLeakage attribute) (CompleteKey attribute) (CompleteProbeMixin attribute) (ContinueSlursReclaimedSlursFull attribute) (Dan_10_0 attribute) (Dan_11_0 attribute) (Dan_6_0 attribute) (Dan_6_2 attribute) (Dan_7_0 attribute) (Dan_8_0 attribute) (Dan_9_0 attribute) (DAN_Jailbreak attribute) (DanInTheWildFull attribute) (Dart attribute) (Davidjl attribute) (Deadnaming attribute) (DiscriminationExclusionToxicityHatefulOffensive attribute) (DRA attribute) (DUDE attribute) (EICAR attribute) (EncodingMixin attribute) (Evasion attribute) (FalseAssertion attribute) (FigStepFull attribute) (FITD attribute) (FunctionMasking attribute) (FutureTenseFull attribute) (GCG attribute) (GCGCached attribute) (GetKey attribute) (GlitchFull attribute) (GOATAttack attribute) (GraphConnectivityFull attribute) (GTphish attribute) (GTUBE attribute) (HF_Files attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HomoglyphObfuscation attribute) (HumanChatbox attribute) (HypotheticalResponse attribute) (InformationHazard attribute) (InjectBase64 attribute) (InjectQP attribute) (InjectUnicodeTagChars attribute) (InjectUnicodeVariantSelectors attribute) (InjectUU attribute) (JavaScript attribute) (JinjaTemplatePythonInjection attribute) (LatentInjectionMixin attribute) (LatentWhoisSnippetFull attribute) (LiteratureCompleteFull attribute) (MaliciousUses attribute) (MarkdownImageExfil attribute) (MarkdownURIImageExfilExtended attribute) (MarkdownURINonImageExfilExtended attribute) (MarkdownXSS attribute) (MisinformationHarms attribute) (MLM attribute) (PAIR attribute) (PastTenseFull attribute) (Payload attribute) (Perl attribute) (PIILeakProbeMixin attribute) (PlaygroundMarkdownExfil attribute) (PrimesFull attribute) (Probe attribute) (Profanity attribute) (Puppetry attribute) (Python attribute) (QuackMedicine attribute) (RakuLand attribute) (Repeat attribute) (RepeatedToken attribute) (RTPBlank attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute) (Ruby attribute) (Rust attribute) (SenatorsFull attribute) (SexualContent attribute) (Sexualisation attribute) (Slurs attribute) (SlurUsage attribute) (SQLInjectionEcho attribute) (SQLInjectionSystem attribute) (STAN attribute) (StringAssemblyDataExfil attribute) (SubFunctions attribute) (Substances attribute) (SystemPromptExtraction attribute) (Tag attribute) (TAP attribute) (TAPCached attribute) (TaskXSS attribute) (Test attribute) (ThreatenJSON attribute) (TopLevel attribute) (Tox attribute) (WhoIsRiley attribute) (Win10 attribute) (WordnetAllowedWords attribute) (WordnetBlockedWords attribute) (WordnetControversial attribute) timeout (InferenceEndpoint attribute) (NVOpenAIChat attribute) to_notes() (AttackState method) top_instructions (FactSnippetMixin attribute) (LatentInjectionReportFull attribute) (LatentInjectionResumeFull attribute) (LatentJailbreakFull attribute) (LatentWhois attribute) (LatentWhoisSnippetFull attribute) (SQLInjectionSystem attribute) TopLevel (class in garak.probes.malwaregen) Tox (class in garak.probes.atkgen) ToxicCommentModel (class in garak.detectors.unsafe_content) Toxicity (class in garak.detectors.perspective) Toxicity_Experimental (class in garak.detectors.perspective) transform() (Base64 method) (Buff method) (CharCode method) (Fast method) (Lowercase method) (LRLBuff method) (PegasusT5 method) TransientConfig (class in garak._config) TranslationMixin (class in garak.probes.latentinjection) TreeSearchProbe (class in garak.probes.base) TriggerListDetector (class in garak.detectors.base) triggers (Davidjl attribute) (Deadnaming attribute) Turn (class in garak.attempt) turns (Conversation attribute)

U

UNLISTED (Tier attribute) Unsubstantial (class in garak.detectors.perspective) untransform() (Buff method) (LRLBuff method) Up (class in garak.detectors.shields) uri (DiscriminationExclusionToxicityHatefulOffensive attribute) (HijackHateHumans attribute) (HijackHateHumansFull attribute) (HijackKillHumans attribute) (HijackKillHumansFull attribute) (HijackLongPrompt attribute) (HijackLongPromptFull attribute) (HumanChatbox attribute) URI (InferenceAPI attribute) uri (InformationHazard attribute) (MaliciousUses attribute) (MisinformationHarms attribute) (RTPFlirtation attribute) (RTPIdentity_Attack attribute) (RTPInsult attribute) (RTPProfanity attribute) (RTPSevere_Toxicity attribute) (RTPSexually_Explicit attribute) (RTPThreat attribute)

URI_ENV_VAR (WatsonXGenerator attribute) use_bytes (EncodingMixin attribute) (InjectSneakyBits attribute) (InjectUnicodeTagChars attribute) (InjectUnicodeVariantSelectors attribute) UUencode() (InjectUU method)

V

valid_format (FileDetector attribute) (PossiblePickleName attribute) verified_results (AttackState attribute)

verify() (AgentBreakerResult method) Vision (class in garak.generators.nim) vulnerability_info (AttackState attribute)

W

WatsonXGenerator (class in garak.generators.watsonx) WebSocketGenerator (class in garak.generators.websocket) WhoIsRiley (class in garak.probes.goodside) Win10 (class in garak.probes.grandma) Win11 (class in garak.probes.grandma)

Win5x5 (class in garak.detectors.productkey) WordnetAllowedWords (class in garak.probes.topic) WordnetBlockedWords (class in garak.probes.topic) WordnetControversial (class in garak.probes.topic) write_report_digest() (in module garak.command)

X

XSS (class in garak.detectors.web_injection)

Z

zalgo() (InjectZalgo method)

ZeroToleranceEvaluator (class in garak.evaluators.base)