60 agents in this category
Runs one request through the imported A2A agents — research briefs, sre diagnoses, developer drafts — and records the run.
Runs one DAC stage against ai-team — the sweep, or one half of the investigation — and publishes what it established.
Sweeps a data cluster's DAC pipeline, investigates the aggregation and settlement halves in parallel when a threshold crosses, and reconciles one verdict on whether eras are being inspected and paid.
Keeps legal, dissolution and entity-formation work current — reads Slack, Notion and mail for anything touching a wind-down track, reconciles it against the documented checklists, and updates them. Use for the daily legal update check, or to ask what changed on a track. (curator)
Diagnostic only: publishes one event and records whether the publish was accepted or rejected, so the vault's answer is visible.
Diagnostic only: answers dac.sweep.requested immediately with a stub verdict, to test whether an agent-published event reaches a subscribed agent.
Voice-first dating coach. Onboarding via conversation; analyses your real dates for what actually mattered.
Coordinates parallel candidate evidence review, deterministic reduction, human decisions and durable adapter handoffs.
Checks supplied candidate proof statements against approved redacted projections and emits evidence only.
Maps supplied, approved candidate projections to role criteria without fetching sources or making hiring decisions.
Agent designed to help users reach specific, measurable objectives, such as completing training, managing mental health, or refining soft skills.
Transcribes audio at a URL with Parakeet TDT and stores the transcript in a cubby.