Writer, developer, Marine Corps veteran. Based in San Diego, building WritersLogic — local-first, cryptographic proof-of-authorship tools for writers — and individually authoring the IETF drafts that formalize the protocol behind them. Contributes to C2PA and W3C standards work, builds open-source AI agent tooling, and competes in applied-ML research (PAN & CLEF). 📍 San Diego, CA
"We Build Proof of Authorship." Local-first, open-standard tools for writers to prove their work is their own:
- WritersProof — desktop app that records typing patterns as evidence of authorship
- Scrivener MCP — open-source AI collaboration tool for Scrivener
- Holographic Memory — privacy-preserving search that processes everything locally
Individually authored, tracked on the IETF Datatracker:
draft-condrey-rats-pop-protocol(-06) — Proof of Process: architecture and evidence formatdraft-condrey-rats-pop-appraisal(-04) — Proof of Process: forensic appraisal and security modeldraft-condrey-posme(CFRG) — Proof of Sequential Memory Executiondraft-condrey-content-binding— content binding for text streams
The PoP drafts extend RATS architecture to bind authorship evidence to document evolution — the protocol layer behind WritersProof above.
Public member, c2pa-org. Submitted PRs (mixed merged/open) to:
- C2PA specifications — search infra, spec fixes
- C2PA soft-binding algorithm list
- W3C AI Content Disclosure CG
cogmem— self-improving, verifiable memory layer for AI coding agentsmisterdev— autonomous LLM build orchestrator: plans a goal into tasks, edits code with anchored search/replace, verifies every change through build/test/lint/typecheck gatescrosstalk— metacognitive multi-agent orchestrator with an adaptive debate swarm, sandboxed verification, and tamper-evident transcripts
Competitive results in AI-text detection and authorship analysis:
trajectory-detection-clef2026— 1st place source detection (0.85 F1), 3rd place safety classificationsimpletext-clef2026— 2nd place, Task 2 (SARI 47.43)voight-kampff-clef2026— AI-generated text detection, 0.891 ROC-AUCstyle-change-clef2026,sense-clef2026— multi-author style analysis, ordinal sensemaking
unicode-safety-check— GitHub Action detecting adversarial Unicode in PRs (invisible chars, bidi attacks, homoglyphs)html-email— zero-dependency, hand-authored HTML email framework, 28 documented client quirks
Not configured — add a classic PAT with repo scope as the TRAFFIC_TOKEN repo secret to enable this.
- Write For Us — 2026-03-30
- Universities Rethinking AI Detection — 2026-03-29
- Freelancers Guide Proving Authorship — 2026-03-14
- Opened PR in dcondrey/buildingforgood
- Released in dcondrey/ai-disclosure-badges
Stars, traffic, blog, and activity sections above update automatically every 6 hours.