Repository Issue Activity (beta)

#2141 Bound generator retries to prevent indefinite hangs

#2157 map euai: tags onto intents

#2136 judge detectors: unparseable judge verdict is scored as a successful attack

#2160 generate() only type-checks _call_model output on 2 of its 4 paths

#884 docs: `Configurable` option descriptions

#1492 probe: conversational `atkgen`

#2155 Buff._derive_new_attempt shares notes and detector_results dicts between attempts, so sibling attempts report the last attempt's scores

#2151 StringDetector normalize (Unicode homoglyph fix, #1884) is never actually enabled — every content-safety detector still bypassable

#1068 Harness that orders probes for expected attack success rate (ASR) according to bag

#973 detector: StrongREJECT -- evaluation for jailbreaks

#1878 probes: shared IntentProbe stub selection within a run

#1971 Your project is on StackMap — a curated map of the AI stack

#2120 agent_breaker detector: non-string success value in judge JSON crashes the whole scan

#2145 SurgeProfanityRacialEthnic loads zero terms — category key mismatch means it never detects anything

#1967 NIM generator aborts entire probe on a single transient HTTP error (408), and reports a misleading error message

#2143 StringDetector matchtype "word" cannot match terms with a non-word character at either edge — 14 shipped Surge terms are dead

#1969 feature: First-class support for Tool messages

#2139 Migrate owasp:llmNN probe tags to year-qualified owasp:YYYY:llmNN format

#2126 Tool Surface Enumeration: Shared Recon Phase for Tool-Aware Probes

#2131 Documentation: REST generator configuration broken in v0.16 — CLI flags changed without docs update

#2110 Optional EvalPort interop for garak attempts and detector results

#87 detector: violence

#2128 elicitation_injection — completion-elicited obfuscated commands (distinct from encoding?)

#2123 feat(cli): add --list_intents to enumerate the intent typology

#2124 intentservice global state (`is_loaded`/`intents_active`) not reset between tests