Looking for workflow testers for the five-flagship Telos stack
Project Telos is now at the point where outside workflow tests are more useful than private polishing.
The current public pieces are:
- gather: source intake receipts
- index: evidence-built workspace maps
- forum: replayable agent-route ledgers
- crucible: measurement-backed MATCH / DRIFT / UNVERIFIABLE verdicts
- telos: shared state, perception, action, and verification loop
What I want to learn: which receipt fields are missing when someone uses this on a real product, research, dev, or agent workflow?
If you have a messy agent run, repo audit, research packet, or product workflow, I would like to test against it and record the pass/fail honestly. Stage: solo, pre-revenue, author-tested, not independently audited.
Main site: https://harperz9.github.io
Field guide: https://harperz9.github.io/field-guide.html
GitHub: https://github.com/HarperZ9
Replies
Concrete test menu if you want a quick way in:
1. Give gather a messy source packet, like a PDF plus a web page plus notes, and check whether source text and derived claims stay separated.
2. Point index at a repo/workspace and tell me which files or dependency edges a maintainer would expect but the map missed.
3. Run forum/telos against a small agent workflow and look for the action receipt a teammate would need to replay what happened.
4. Give crucible a claim with an intentionally missing source and see whether it returns UNVERIFIABLE instead of laundering the gap into confidence.
The most useful reply is a failure case: "I could not trust this in my workflow because receipt field X was missing."