Macro is a unified workspace for teams: email, chat, docs, tasks, agents, calls, and CRM — @-linked together with shared AI memory.
TypeScript agentaiai-agentsall-in-onecrmcrm-systememaillinearmcpmessagingnotesnotionofficerustslackslack-alternativestartupstartup-toolstypescriptworkspace
import '../../tooling/just/database.just'
# Run the real task-dedup semantic eval (the opt-in #[ignore] tests): real
# OpenAI embeddings + Anthropic judge over the labeled corpus, reporting
# precision/recall, retrieval recall@k, the vector-score threshold sweep, and
# judge variance. Needs the local macrodb up and OPENAI_API_KEY /
# ANTHROPIC_API_KEY / CEREBRAS_API_KEY in the environment. Every file in
# tests/eval/cases/ is discovered automatically (see tests/eval/main.rs). Each
# measurement runs its model calls concurrently, and the measurements run on
# parallel test threads, for the fastest wall-clock; pass a name to run just
# one (e.g. `just eval judge_accuracy`).
eval filter="":
DATABASE_URL={{ DATABASE_URL }} cargo test -p task_dedup --test eval {{ filter }} -- --ignored --nocapture
test:
DATABASE_URL={{ DATABASE_URL }} cargo test