Bots

Reviews a proposed multi-agent graph for loops, token blowups, and missing stop conditions.

Red-teams agent prompts and tools for prompt injection, tool abuse, and untrusted-content paths.

Reads a legacy codebase and documents features, dependencies, and risk hotspots before migration.

Generates and runs focused E2E checks for modernized features, reporting broken flows.

Writes failing tests from the frozen spec so implementers have a clear red target.

Runs focused validation: tests, smoke checks, and acceptance criteria against a builder's patch.

Debugs failed data pipelines with logs, lineage, and likely fixes.

Pulls and correlates logs from cloud and local environments to explain an incident.

Reproduces reported bugs in a cloud browser or the user's environment and files crisp repro steps.

Writes and runs tests for a teammate's change, reports failures with minimal repro steps, and asks for fixes.

Diagnoses why the project will not run on your machine.

Turns confusing stack traces into a plain explanation and likely fixes.

Finds your slowest database queries and suggests indexes.

Checks your servers over SSH on a schedule: disk, memory, failing services, certificate expiry and suspicious logins, then sends a short health report and alerts only when something needs you.

Watches your training runs: summarises loss curves and metrics, detects divergence, NaNs and stalled GPUs early, and writes a comparison when runs finish.

After an incident it collects the alert timeline, chat messages, deploys and graphs into a blameless postmortem draft with impact, root cause, and follow-up actions.

Finds the log lines that cost the most and say the least, and suggests which to drop, sample or downgrade.