v0.1.22

What's Changed

• docs(routing): route to skills, not only role agents (N9) by @AlexShchuka in #69

Full Changelog: v0.1.21...v0.1.22

v0.1.21

What's Changed

• feat(eval): vanilla/no-plugin baseline arm by @AlexShchuka in #68

Full Changelog: v0.1.20...v0.1.21

v0.1.20

What's Changed

• feat(eval): cross-family second judge via GitHub Models (free, models:read) by @AlexShchuka in #67

Full Changelog: v0.1.19...v0.1.20

v0.1.19

What's Changed

• fix(eval): make Krippendorff α rater-aware — runs were masquerading as raters by @AlexShchuka in #66

Full Changelog: v0.1.18...v0.1.19

v0.1.18

What's Changed

• fix(eval): correct inverted McNemar tail — gate was passing regressions, blocking improvements by @AlexShchuka in #64
• fix(eval): zero-variance paired diffs are infinite effect size, not zero by @AlexShchuka in #65

Full Changelog: v0.1.17...v0.1.18

v0.1.17

What's Changed

• add skills: agent-fleet-orchestration, adversarial-review, robust-by-construction by @AlexShchuka in #62

Full Changelog: v0.1.16...v0.1.17

v0.1.16

What's Changed

• feat(skills): два первых скилла плагина — paper-to-code + harness-improve by @AlexShchuka in #61

Full Changelog: v0.1.15...v0.1.16

v0.1.15

What's Changed

• feat(hooks): issue #55 п.2+п.4 — human gate для протокольных артефактов, обогащение cycle-trail (Phase 0) by @AlexShchuka in #59

Full Changelog: v0.1.14...v0.1.15

v0.1.14

What's Changed

• calib(protocol): issue #55 п.1–3 — три инварианта, матрица клейм-тайпинга, рубрика systems-thinking by @AlexShchuka in #58

Full Changelog: v0.1.13...v0.1.14

v0.1.13

What's Changed

• docs(roadmap): N2 shipped — positive eval probes q01–q06 landed (PR #34) by @AlexShchuka in #57

Full Changelog: v0.1.12...v0.1.13