Agent fleets with a replayable causal ledger.

ZentropyLabs.ai - I build systems for tomorrow.

by•2mo ago

Systems engineer and cross-domain builder whose work spans AI-agent infrastructure, compilers and language tooling, real-time graphics and color science, verification systems, docs, and operations. Builds inspectable tools with explicit boundaries, tests, documentation, provenance, and reproducible release paths. Independent engineering practice, grounded in earlier technical support, freelance technical writing, compliance documentation, customer-facing operations, and field leadership.

Replies

Best

Hunter

📌

I built Project Telos because model self-reports are not enough for real work. The line is built around checkable state: witnessed intake, rerunnable workspace maps, multi-agent ledgers, measured judgment, and a shared human-AI surface. The honest stage: solo, independent, pre-revenue, and pre-proof on the largest thesis. I am looking for verification and testing against real workflows, technical pushback, early traction from people willing to inspect the receipts, and possibly modest grassroots research funding. Main site: https://harperz9.github.io GitHub: https://github.com/HarperZ9 Flagship repos: https://github.com/HarperZ9/gather, https://github.com/HarperZ9/index, https://github.com/HarperZ9/forum, https://github.com/HarperZ9/cruc..., https://github.com/HarperZ9/telos

Report

2mo ago

Hunter

I added a field guide for testing Project Telos on real workflows:

https://harperz9.github.io/field-guide.html

The guide names five first test environments: doctor/admin, artist/studio, media/newsroom, token economy/routing, and reasoning. The ask is not applause. I am looking for people willing to verify the tools, run them against real workflows, find where the receipts break, create early traction, and point toward modest grassroots research funding if the work should keep moving.

Public repos:

gather - https://github.com/HarperZ9/gather

index - https://github.com/HarperZ9/index

forum - https://github.com/HarperZ9/forum

crucible - https://github.com/HarperZ9/crucible

telos engine - https://github.com/HarperZ9/telos

Report

2mo ago

Hunter

If you want to test it today, the shortest path is the field guide: https://harperz9.github.io/field-guide.html

I would especially value testing in five concrete environments:

1. Doctor/admin handoff: can a messy intake become a witnessed digest that a reviewer can re-check instead of trusting the model summary?

2. Artist/studio: can a person and model work the same artifact while keeping seed, source, and export receipts visible?

3. Media/newsroom: can a public claim or asset carry a re-checkable trail instead of just a confidence score?

4. Token economy/routing: can an agent route work using a compact map instead of rereading the whole workspace every turn?

5. Reasoning: can a judgment loop say MATCH / DRIFT / UNVERIFIABLE, then refine the weakest axis without pretending certainty?

Main site: https://harperz9.github.io

GitHub: https://github.com/HarperZ9

Flagship repos: https://github.com/HarperZ9/gather · https://github.com/HarperZ9/index · https://github.com/HarperZ9/forum · https://github.com/HarperZ9/crucible · https://github.com/HarperZ9/telos

I am looking for verification, testing, traction from real users, and possibly modest grassroots research funding or pointers.

Report

2mo ago