Skip to content

USERS.md

Latest commit

 

History

History
122 lines (86 loc) · 4.88 KB

USERS.md

File metadata and controls

122 lines (86 loc) · 4.88 KB
Raw

Users and Community

The current source version is v6.3.0, and the latest published release is v6.3.0.

Track record: the honest version

Godpowers has zero recorded production users.

That sentence stays at the top of this page until it is no longer true. A tool whose entire purpose is refusing to let software claim more than it can prove does not get to make an exception for its own marketing.

Here is what actually exists today. Godpowers has a large, tested surface: an optional read-only companion package, executable stage gates, three host-run proof studies, three externally verifiable adoption canaries against real third-party repositories, and an accountability system that has been hardened against its own failure modes. All of that is real and all of it is checked by tests.

What none of it is: evidence that somebody shipped a product with this and it worked. Real users are the only thing that reveals which gaps actually matter, and we do not have them yet.

  • [DECISION] The Phase 2 host proof campaign completed Run A for local and CI-verifiable CLI proof, Run B for local web-app proof, and Run C as a blocked-but-documented host proof.
  • [DECISION] Slot A does not prove production usage, deployed smoke, or token-dollar accounting because no staging origin or cost.recorded events were captured.
  • [DECISION] Current evidence and blockers are recorded in Run A, Run B, and Run C.

Note that Run C was blocked and we published it anyway. A case study collection containing only successes is a marketing brochure, not evidence.

The proof we actually need

The next credibility milestone is not a bigger command count. It is one real /god-mode run, on a codebase nobody here knows, that produces work which ships or is genuinely ready to ship.

Before anyone claims a case study, run all three of these and record whether the shipped fixture suite still passes on the local host:

npx godpowers quick-proof --project=.
npx godpowers dogfood
node scripts/run-adoption-canary.js <git-url>

Then write down all of it, including the parts that went badly:

  • The repository shape and project type
  • Wall-clock time
  • Token and dollar cost from /god-cost
  • How many times it paused, and why each time
  • Assumptions that turned out wrong, and what it took to repair them
  • Which validation commands ran, and what they returned
  • The host guarantee level from /god-status
  • Quick proof, dogfood, and adoption canary results
  • What actually shipped, or precisely what blocked it

Using it on something real?

Then you are ahead of us, and we want to hear about it:

  1. Open an issue describing what worked and what did not. The second half is more useful than the first.
  2. Run /god-extract-learnings after each milestone, and consider sharing what it captured.
  3. Tell us about workflows that should exist and do not.

Where to find us

Channel For
GitHub Issues Bug reports and feature requests
GitHub Discussions Questions and sharing experiences
Discord Realtime chat (coming soon)

Case studies

External CLI canaries, run against real cloned repositories:

Host-run proof studies:

What these do and do not establish:

  • [DECISION] The CLI canaries are not production-user studies.
  • [DECISION] The CLI canaries prove first-contact CLI signals against real cloned repositories and keep narrow proof limits explicit.
  • [DECISION] The Codex host studies are host-run evidence, but they are still not production user adoption.

If you ship something with Godpowers, write it up. We will feature you here.

Milestones we are watching for

None of these has happened yet. Each one would be a first:

  • A production deployment using /god-mode
  • An incident handled with /god-hotfix, followed by /god-postmortem
  • A completed /god-upgrade migration
  • A skill pack contributed by the community
  • A team using /god-sprint to set their cadence
  • An external extension pack scaffolded, tested, and published
  • A multi-repo suite release dry-run validated against real dependent repos

Using it alongside other tools

If you have combined Godpowers with another AI coding workflow system, tell us what worked, where the two collided, and how you resolved it. We document these patterns in references/shared/ORCHESTRATORS.md.

License

MIT. See LICENSE.

Code of Conduct

See CONTRIBUTING.md.