Skip to content

docs(competitor-watch): protocol spec — in-repo audit trail - #34

Merged
gHashTag merged 1 commit into
mainfrom
feat/competitor-watch-spec
Jul 4, 2026
Merged

gHashTag merged 1 commit into
mainfrom
feat/competitor-watch-spec

Conversation

@gHashTag

@gHashTag gHashTag commented Jul 4, 2026

Copy link
Copy Markdown
Owner

Summary

Publishes the competitor-watch protocol in the repo so a third party can replicate the weekly monitoring without access to the platform scheduler. Keeps the tri-net auditability thesis honest — the monitor that produces claims about competitors cannot itself be a black box.

What is in the spec

  • Ten product-name queries per week (not company names, to keep M&A / earnings noise out).
  • Four arXiv keyword sweeps.
  • Six relevance triggers and an explicit exclusion list.
  • Primary-source discipline (Reddit / X / LinkedIn never cited directly).
  • Draft-PR-only filing procedure. Silent on empty weeks.
  • Dedup rule against the previous 7 days.
  • Hand-reproducibility recipe.
  • Future-work note on migrating the executor to .github/workflows/competitor-watch.yml once search-API secrets are provisioned.

Executor (implementation detail, not covered by this PR)

Current executor is Perplexity platform cron 64822c1c, cadence 0 2 * * 5 UTC = Fri 09:00 Asia/Bangkok. It runs the protocol above. If the two ever disagree, this spec wins and the executor gets fixed. Migration to a fully in-repo GitHub Actions workflow is called out as future work and needs search-API secrets before it can land.

Non-goals

  • No CI wiring.
  • No metrics, no benchmarks — this file only defines what to look for.
  • No change to any existing module.

Anchor: phi^2 + phi^-2 = 3

Publishes the competitor-watch protocol in the repo so any third party can
replicate the weekly monitoring without access to the executor. Keeps the
tri-net thesis honest: a monitor that lives only inside a proprietary
scheduler is the black-box anti-pattern the delta paper critiques. The
scheduler is now an implementation detail; the protocol is portable.

Contents:
- Ten product-name queries (not company names, to keep M&A noise out).
- Four arXiv keyword sweeps.
- Six relevance triggers + explicit exclusion list.
- Primary-source discipline (Reddit / X / LinkedIn never cited directly).
- Draft-PR-only filing procedure, silent on empty weeks.
- Dedup rule against the previous 7 days.
- Hand-reproducibility recipe.
- Future-work note on migrating to a GHA workflow once search-API
  secrets are provisioned.

Current executor: Perplexity platform cron id 64822c1c, cadence
0 2 * * 5 UTC (Friday 09:00 Asia/Bangkok). If the executor and this spec
disagree, the spec wins and the executor gets fixed.

Anchor: phi^2 + phi^-2 = 3
@gHashTag gHashTag added mesh TRI-NET mesh track documentation Docs labels Jul 4, 2026
@gHashTag
gHashTag marked this pull request as ready for review July 4, 2026 13:06
@gHashTag
gHashTag merged commit 91a5b63 into main Jul 4, 2026
2 checks passed
@gHashTag
gHashTag deleted the feat/competitor-watch-spec branch July 4, 2026 13:06
gHashTag pushed a commit that referenced this pull request Jul 4, 2026
…dit-trail primitive

Bounded expansion (not full v1). The skeleton previously proposed the
auditability primitive abstractly (Sections 1-4); this commit adds Section
5 reporting that the primitive is now empirically realized end-to-end at
tri-net main dc1bebb + t27 master 3c912d9.

Section 5 subsections:

- 5.1 What is materialized -- one spec (specs/wire.t27), three generated
  backends (gen/{rust,zig,c}/wire.*), one workflow (spec-drift-guard.yml),
  consumer path under the same umbrella (src/wire.rs -> u32_be).
- 5.2 Audit-trail tuple (concrete instance) -- (tri-net@dc1bebb,
  t27@3c912d9, spec-drift-guard run) is the minimal fetch set for a third
  party to independently verify byte-identity.
- 5.3 Merge chain table -- 5 tri-net PRs (#33/#34/#35/#37/#38) + 2 t27
  PRs (#1320/#1337) with merge SHAs.
- 5.4 Empirical checks at HEAD -- three diff -u results (all empty), zero
  'unsupported: ExprCast' in Zig/C output, 101 lib tests green including
  header_roundtrips as empirical byte-order-equivalence proof.
- 5.5 What this reference impl does NOT show -- byte-identity is not
  functional correctness; only one spec is covered so far; no silicon;
  trust in t27c itself is not eliminated (moved, per Carrone 2026).

Sync updates across the doc:

- Header cross-link: 91a5b63 -> dc1bebb (post-#38).
- Section 3.1: describes all three gen targets, not just Rust.
- Section 3.2: drift-guard covers Rust + Zig + C; links to #35 + #38.
- Section 3.3: ExprCast marked resolved on all four backends
  (Rust #1320, Zig+C #1337, Verilog pre-#1320).
- Section 6 (Limits): removed obsolete 'Rust-only SSOT until #1333' bullet;
  added trust-surface-of-t27c and no-HDL-target-yet bullets.
- Section 7 (Future work): struck through completed Zig+C drift-guard
  item; added protocol-stack expansion, t27c deterministic-build items.
- Appendix A: three-target reproducibility recipe (8 steps), pins
  tri-net@dc1bebb + t27@3c912d9.

Draft only -- not for external circulation.

Anchor: phi^2 + phi^-2 = 3.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

documentation Docs mesh TRI-NET mesh track

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant