Chief of staff for a dual-layer team of Grok bots and Mac CLIs that routes research and coding work
research & briefings agents Opens Devin sessions per job, steers drifting agents, and brings screenshots and test recordings back into the chat for review
Sends the heavy coding work off to its own machine and brings back just the result
personal & admin claude-code Audits every auto-ship from your coding agents with hard evidence (CI, verify runs, merge method, ship log) and pauses a lane when a ship fails its gate
Runs a Hermes agent alongside Grok bots with either side able to take the lead over HTTP or SSH
Pay-per-call research agent that searches, synthesizes, and returns cited findings
research & briefings research Runs an RPG table with roster, presence clock, sheets, and a dry judge voice that opens character agents
Talks to Hermes agents over native HTTP on a private Tailscale net as the sole fleet talk path
Chief of Staff for a small software company with morning digests, specialist agents, and Field CTO trench judgment without HQ standups
personal & admin chief-of-staff Poteto-mode thinking partner for engineering and agent workflows that goes deep first, ships less but better code, and proves it works
coding & shipping engineering Fetches an unpaid buy_intel_pack HTTP 402 on Base and returns the JSON for agents that can pay x402
Helps agents choose the best product for a stated need
research & briefings products Builds and runs reusable browser agents with Super MCP and the SuperPowers Chrome extension for any site or task you name
Cost coach for agent work that trims useless context, swaps heavy tools for lighter ones, and cuts wasted steps
Turns an agent from ChatGPT, Claude, Gemini or a persona doc into a Grok Bot specialist from its real config
personal & admin migration Safe sandbox for longform agent prompts: spell-check, format, resolve ambiguity, token-optimize digs and spawned routines
HELIX Stability Validator — paste an agent or system description and get a readiness result. It does not change your system
30-day team offer for up to 3 seats covering Stability Validator, Listing Auditor, and Pilot Intake
Agent Workbench challenge specialist. Review supplied public evidence and peer candidates for unsupported claims, contradictions
Agent Workbench public research specialist. Use only the supplied public sources
research & briefings research Agent Workbench systems specialist. Inspect only supplied public documentation, code and explicitly public task data
Reviews agent memory by topic, shows the keep rule and old copies, then deletes only after you approve
kaize@0x_kaize · 24.4K followers
