Skip to content

docs: miner guide — hosted quickstart, attested-execution direction, broadcast-vs-fetch model#331

Draft
wallscaler wants to merge 1 commit into
mainfrom
claude/gallant-carson-7Az6m
Draft

docs: miner guide — hosted quickstart, attested-execution direction, broadcast-vs-fetch model#331
wallscaler wants to merge 1 commit into
mainfrom
claude/gallant-carson-7Az6m

Conversation

@wallscaler

Copy link
Copy Markdown
Contributor

What

Three small, customer-facing edits to docs/FAST_PATH_MINER_GUIDE.md:

  1. Quickstart with two honest paths. "Fastest: hosted on Polaris (pre-wired, minutes to first solve)" and "Manual: any box." Explicitly states both paths earn identically and nothing in the protocol is provider-specific — the golden path is a convenience, not a gate.
  2. Solver provenance: where it's going. One paragraph stating the direction already implied by the announcement ("attested execution is coming"): TDX/DCAP quotes binding solver_hash to the run, any provider's valid attestation counts, attested runs to earn preferential credit. Hosted providers named as the turnkey route for miners who don't want to wire DCAP themselves.
  3. The read model, stated plainly. Adds the one-liner from the launch discussion: the broadcast tells you when, one signed fetch gets yours — the phrasing that stops "get get get" loop-polling while explaining why the challenge fetch itself is per-hotkey.

Also fixes the broken #solver-provenance anchor (heading slug is #3-solver-provenance).

Why

  • Time-to-first-solve is the adoption metric for the just-announced fast path; the guide's first screen should offer the shortest path without hiding the manual one.
  • Miners are being asked to send solver_id/solver_hash today on trust; the guide should say what that buys them tomorrow, and that the attestation rules will be provider-neutral.

Notes

Docs only — no code paths touched. Follows up on #328/#329.


Generated by Claude Code

…broadcast-vs-fetch model

- Quickstart with two honest paths: hosted (Polaris, pre-wired) and manual
  (any box); both earn identically, protocol stays provider-neutral.
- Solver provenance: state where it goes — attested execution (TDX/DCAP),
  any provider's valid attestation counts, attested runs to earn
  preferential credit.
- Make the read model explicit: the broadcast tells you when, one signed
  fetch gets yours.
- Fix the solver-provenance anchor link.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant