DUSK v0.2.0

What's Changed

• feat: Flask API, all 6 partner integrations, self-healing, enterprise demo by @TFT444 in #14
• merge: bring all feature branches into main by @TFT444 in #15
• feat: Vercel deployment config by @TFT444 in #16
• fix: resolve Vercel 404 -- bundle demo/src files, fix path by @TFT444 in #17
• docs: add team section with all contributors by @TFT444 in #18
• docs: update contributor name to He Xiao by @TFT444 in #19
• fix: seed demo decisions so dashboard shows live data on cold start by @TFT444 in #20
• fix: rich seed data -- 15 realistic decisions across all verdict types by @TFT444 in #21
• feat: Inside DUSK architecture page with live backend stats by @TFT444 in #22
• Gate service: /v1/gate HTTP endpoint with SIE encode/score/extract (R0-R8) by @ritiksah141 in #36
• chore: remove Vercel deployment config by @TFT444 in #38
• chore: split CODEOWNERS between owner and co-owner lanes by @TFT444 in #39
• feat: agent-demo/mock-prod foundation (T0, T1, T2, T5) by @TFT444 in #53
• feat: land remaining agent-demo pipeline on dev (T3, T4, T6, T7, T8, T9, T10) by @TFT444 in #56
• Gate service R9-R12: integration, real load test, hackathon cleanup, SIE gate hardening by @ritiksah141 in #57
• feat: make examples/agent-action-monitor self-contained (closes #58) by @TFT444 in #59
• fix: agent-action-monitor README/config bugs from pre-submission review by @TFT444 in #61
• fix: config doc accuracy, stable fallback hash, bounded webhook threads, Python version docs by @TFT444 in #71
• fix: TraceDecision -- prune dead fields, reconcile risk_flags and similar_decision_ids by @TFT444 in #72
• ci: add lint/typecheck/security/test jobs for examples/agent-action-monitor/ by @TFT444 in #73
• fix: recursively flatten nested change payloads for anomaly scoring by @TFT444 in #74
• fix: heal.py -- real self-healing, wired into dusk gate --heal by @TFT444 in #75
• ci: dead-code detection, secret scanning, doc-consistency check by @TFT444 in #77
• feat: enterprise-grade repeat-offense memory for the live gate by @TFT444 in #80
• fix: derive similar-decision verdict from the real recorded verdict, not score by @TFT444 in #79
• fix: background persistence and safe default path for OffenseMemory by @ritiksah141 in #84
• docs: match superlinked/sie gallery README conventions by @ritiksah141 in #83
• docs: polish example comments and docstrings by @ritiksah141 in #86
• docs: polish root comments and docstrings by @ritiksah141 in #88
• fix: bound n8n webhook queue and add decision-history fairness by @ritiksah141 in #90
• chore: remove dead Aikido/Mubit references and root TraceDecision's dead fields by @ritiksah141 in #92
• docs: polish Superlinked package and root architecture by @ritiksah141 in #94
• docs: simplify README hero banner by @ritiksah141 in #95
• Dev by @TFT444 in #96
• docs: visualize network sweep detection flow by @ritiksah141 in #98
• docs: sync latest dev documentation to main by @ritiksah141 in #99
• chore: prepare OWASP Incubator readiness by @ritiksah141 in #106
• docs: finalize OWASP application submission package by @ritiksah141 in #108
• ci: add fast security and sandbox gates by @TFT444 in #112
• feat(policy): add enterprise agent-action rule pack by @TFT444 in #115
• feat: add reproducible OWASP application demo by @ritiksah141 in #111
• release: promote DUSK v0.2.0 OWASP Incubator candidate by @TFT444 in #116
• fix(security): strict bool coercion, auth fail-closed, stable health codes by @TFT444 in #123
• chore(release): prepare OWASP v0.2.0 evidence by @TFT444 in #124
• release: promote v0.2.0 readiness to main by @TFT444 in #125

Full Changelog: https://github.com/ShieldTech-Ltd/DUSK/commits/v0.2.0