One studio, two threads: products that tell defenders what adversaries do next, and tools that hold writing to a human standard. Here's what we run.
Forward calls, source triage, and the engines underneath them.
A paid weekly wire of ten forward calls on what attackers do next, each scored and mapped to your priority intelligence requirements.
Read a sample →LLM-driven builder for security-vendor integrations. Describe the vendor, get a working MCP connector instead of a week of glue code.
connector-forge.com →Generate cyber tabletop exercises for SOC and IR teams: scenario, injects, and facilitator notes, built from real adversary behavior.
Preview →Every morning, the firehose narrowed to the ten items that actually touch your assets. The triage layer beneath the wire.
Preview →Rewrite a threat report for the audience that has to act on it: SOC, CISO, exec, or board. One report, four altitudes.
PyPI →The shared engines: score relevance against priority requirements, and lint analytic writing for missing confidence, overclaims, and unsourced lines.
GitHub →Serial fiction, and the anti-slop tech that keeps machine writing in a human register.
A reading and writing platform for serial fiction, with an anti-slop gate on every chapter and readers who investigate the world, not just consume it.
Start reading →Social posts and replies in your own voice, learned from how you actually write. Voice-DNA in, slop out.
Preview →An editor extension that flags the tells of machine writing as you type. The validator behind the Voice-DNA work.
Open VSX →