agent-workflow

Explain using pragmatics and semantics, no syntax please.

Stress-test any artifact by spawning an Adversary agent (finds problems) then a Counterweight agent (calibrates findings). Use when the user says "adversarial review", "challenge this", "stress test this", "red team this", or wants a second opinion on a design or implementation.

Run a self-perpetuating, multi-agent improvement loop on ANY codebase (chat, ingest, a UI, infra — anything with tests). A committed backlog of file-disjoint, waved tasks is implemented in parallel by a background Workflow, verified, adversarially reviewed, fixed, committed, and then re-scouted into the next round's backlog. Use when the user wants to iteratively harden/improve a project across many tasks ("improve X", "run a round", "keep iterating on Y", "set up an improvement loop"), to continue an existing loop ("run the next round"), or to fold a fresh ask into the next round. Requires the Workflow tool (multi-agent orchestration).

Schedule a delayed prompt that fires in the current Claude Code session at a target time, by launching a background bash task that exits at the target. Use when the user says "later", "in N hours/days", "wake me", "remind me in", "tomorrow", "next week", "schedule a follow-up", or wants to defer work past the 1-hour /loop ceiling. The wake payload can be a reminder, an instruction to execute, or a slash command.

Create a background polling loop that monitors an external system and wakes Claude only when something changes. Use when the user says "monitor", "watch", "keep an eye on", "poll", "wait for", or wants to be notified when something external finishes or changes (CI, deployments, build jobs, API status, review approvals, etc.).