v0.19.3

What's Changed

• fix: guide agents past iOS keyboard dismissal and get-text guesses by @thymikee in #1173
• fix: make settle tail list real actionable targets by @thymikee in #1172
• feat: fix codex runner, add --override-doc grading, port skillgym quiz cases to help-conformance bench by @thymikee in #1176
• fix: add curated suggestions for open-url and close-session by @thymikee in #1175
• docs: ADR for interactive replay and resolution disclosure by @thymikee in #1177
• chore: add agent-efficiency regression guards by @devin-ai-integration[bot] in #1174
• refactor(daemon): route perf metrics sampling body through PlatformPlugin facet by @devin-ai-integration[bot] in #1191
• test(output-economy): add routine-workflow output-behavior oracle by @devin-ai-integration[bot] in #1190
• fix: align Maestro swipe semantics across platforms by @thymikee in #1179
• ci: ratchet against production-unused exports by @thymikee in #1202
• feat: add derived fail-open check:affected selector by @devin-ai-integration[bot] in #1195
• refactor: remove production-unused exports by @thymikee in #1203

Full Changelog: v0.19.2...v0.19.3