the project lead's 6 October 2026 ask: deep backward and forward research across the hash, finality,
economy, network and every shipped surface. This commit carries the first three lanes.
- docs/analysis/horizon/algorithm.md: the chip model on the 6 October numbers (f = 1 GDDR7
chip 5.7x per joule against the 5090 at class v3, 2.1x at class v4 with k = 1), the FPGA
lane tightened to 0.30x to 0.47x per watt, the reserve R0 to R8, the reconciled shadow-N
ladder (section 5.3a) with HBM4 and three verifier brackets, the first measured verifier
proxy on igneum-build-1 (class v4 5.06 ms cold, dr736 10.51: out), the dataset schedule
to 2030; model sim/horizon/algorithm/model.py.
- docs/analysis/horizon/frontier.md: sixteen ideas ranked by payoff over difficulty with the
Monero and Kaspa attacks, prior art cited, the honest never column; model
sim/horizon/frontier/frontier_model.py.
- docs/analysis/horizon/new-pow.md sections 0 to 4: three new proof-of-work schemes defined,
reviewed in two personas, scheme A (mining is proving) ruled out on bytes and
sampleability, B and C in prototype on two rented 4090s; measured rows follow.
- docs/analysis/horizon-2026-10.md: the summary skeleton and the lane table.
Every rental cost cites docs/bench-log.md "Rental cost of hash, 6 October 2026".
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
From the Horizon lane analyses of 6 October 2026 (docs/analysis/horizon/algorithm.md sections 5.1, 5.4 and 8;
docs/analysis/horizon/frontier.md sections 3.11, 4.1 and item I13; both land with the lane's own commit).
(1) Wherever the litepaper or the home page implied that the hourly program, the era draw or the instruction reserve
defeat a chip by surprise (the hero SVG line, the home hourly-program note, the Mining section's three ideas, the
"Every six months" row, the "A chip is impossible" item), the text now says what holds: they are automatic schedule
changes against fixed datapaths and against human forks; a chip wired for one program is useless; against the chip
that stores the dataset every drawn parameter is firmware and everything it needs is public at genesis, so the defence
is the latency-shadow work (class v4) and the price per joule. Ledger M32.
(2) docs/analysis/chip-model-v3.md section 5.3: the HBM activate-bound ceiling (8 per 12 ns, 10.7 G reads/s a stack)
is marked UNMEASURED beside the JEDEC HBM2 figure (tFAW 28 ns, 4 activates: 2.3 G), and the public FPGA line carries
only the measured row (Shuhai, FCCM 2020: 2.4 G reads/s, 0.30x to 0.39x of the RTX 5090 per watt) until an AWS F2
hour measures the ceiling. Ledger M33.
(3) The finality section's "What is not here" paragraph and the glance table's Finality row carry, verbatim: "No coin
is staked. The only thing at stake is 30 days of public work: a vote key's weight is its blue blocks over the window,
and equivocation strips it for 30 days." Ledger F26.
(4) "For miners", under the three-streams table: all of Ethereum L1's proving is about USD 36 a day at the September
2026 tracker cost (a secondary source) against about USD 13,700 a day of year-1 emission at USD 0.005 per IGN (the
price an input, not a forecast), so external proving is a small second income at launch and the lottery pays the
bills. Ledger E19.
docs/fud-ledger.md gains the four rows (Conceded, stated, 6 October 2026); site/ledger.html regenerated (171 entries);
tools/ci/ledger-text-check.mjs carries the five new stated sentences (48 sentences, 0 missing).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The rebase onto e69117a kept master's captions and build scrub and this branch's wheel handler, chip and note. The build
re-inlined the journey feed from master's journey.json. A code comment in site/live-dag.js named the owner; master's
identity grep now covers every served file and caught it, so the comment says the owner.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
6 October 2026: from a release worktree, `git push origin master` pushed the shared checkout's unchanged local master ref and the tool reported "pushed to origin/master" from its own arithmetic; the Windows build was then looked for on master. Master merges are main's.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
the project lead, 6 October 2026: the miners table and the events feed were too long for the story. The miners card is now a strip:
the count ("66 vote keys in 10 min"), the top five lanes by blocks with a bar each and the last-seen time, and "and N more"
with the bench table linked. The events card shows the last five, one line each, with the count of the rest in the note.
Two columns from 900 px. Measured headless: at 1440 both cards end at 1.9 screens (the scene fills the first); at 390 the
miners card ends at 2.5 screens and events at 2.95; no horizontal overflow, no console errors. Captures replaced in
docs/plans/site-ui-3-shots/after/live-1440-dark.jpg and live-390-dark.jpg.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
the project lead, 6 October 2026: scrolling past the DAG zoomed it. The module's wheel listener took every vertical wheel. Now a plain wheel
or a trackpad's two-finger scroll is never captured: the handler returns before preventDefault unless ctrl or cmd is held or
the scene is engaged. A click into the scene engages it (the canvas takes an ember outline and a chip reads "scroll to zoom,
Esc to release"); Escape, a click outside the scene or the chip releases it. Pinch zoom is unchanged. Same module on / and
/live; both pages carry the chip and the note says how to zoom. opts.chip and opts.onEngage are new, optional; dag.engage(bool)
is exposed.
Proved headless on both pages: a plain wheel over the scene scrolled the page 300 px and left the window at 120 s; engaged, a
wheel zoomed the window from 120 to 138 s with the default prevented; the chip showed on click and hid on Escape; no console
errors. Link check 884 links 0 broken.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- infra/build-server/night/night-battery.sh: checkout of master and the newest release-*-node fork branch under
/srv/builds/_night, cargo test --release --no-fail-fast per crate (repo and fork), igneum-pow fuzz at 10x, the three
simulators in full, the fast-time harnesses (finality-attacks, harness s3 s4, exec-sync reorg) on binaries built into
target-integration, clippy per crate dir, cargo audit per Cargo.lock; docs/benchmarks/night/<date>.md with a pass/fail
table and "new since last night"; committed as igneum-labs on night-battery and pushed to the mirror, never master.
NIGHT_SUBSET=1 is the dry-run subset. The unit runs it through remote-run.sh so the slot spans the invocation.
- igneum-night-battery.{service,timer}: 02:00 Europe/London, User build, Nice 19, idle IO, 8 h limit, not Persistent.
- provision.sh: step_cargo_tools (cargo-audit), step_night (files from the mirror at NIGHT_REF, timer enabled), innoextract.
- tools/repro/rebuild-release.sh + infra/build-server/repro/rebuild-on-box.sh: pins from docs/plans/release-<v>.md
("(node <sha>, app <sha>)"), shipped hashes from the public downloads (the HiveOS tarball, the installer via
innoextract) or --shipped, a clean clone of repo and fork on the box, two passes per target without sccache under build
slots, MATCH or DIFFER per artefact against the shipped bytes and against the other pass, with the reason read off the
binaries (glibc version needed, PE timestamp, commit string); evidence into docs/evidence/reproduced/<version>.md.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Copy, from origin/master 0a63474 (site audit). The home and miner pages named "PC 1" in the figure captions, the
page-by-page lead, two alt texts, the Ember Tune table title, the Ember row's source line and the home pill; they now
say "a Windows rig with an RTX 5090, RTX 4070 and RX 9070 XT" at the first mention on each page and "the Windows rig"
after. The generated pages carried the same names from the bench-log (80 on /bench, 7 on /evidence, the inlined
journey on the home page): site/scrub.mjs now maps PC 1 to "the three-card Windows rig (RTX 5090, RTX 4070, RX 9070
XT)" and PC 2 to "the RTX 5090 Windows rig", build.mjs re-scrubs the stored journey entries, two /bench anchors on the
miner page follow the renamed headings, and \bPC [12]\b joins site/forbidden-strings.txt so the build fails if a number
returns. The scrub also covers the audit's other page-leak shapes (a pid, 0.0.0.0:port, ~/.config paths, --rpclisten=),
which the bench page carried and which now fail the build if they return.
CI: tools/ci/forbidden-strings.txt's appended audit block sat on one physical line with literal \n text, so none of its
patterns was active; \bPC [12]\b is now a real line there and the identity check's export scrub maps the two machines
the same way (igneum-public/tools/sync.sh must carry the same two rules). The other four audit patterns moved to
site/forbidden-strings.txt, since the public export carries simulator schedule logs where a pid is a pid. The identity
check gains a second pass over every served file under site/ (html, json, txt, xml, webmanifest, css, js; not api/ or
the build scripts), unscrubbed, with both pattern lists; dl\.igneum is narrowed to the tokened path so the public
download buttons pass. Shown to fire on a page naming PC 1 (exit 1) and to pass on the tree (0 hits over 232 export
files and 32 served site files).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- provision.sh step_runner: actions/runner 2.338.0 (sha256 checked) at /opt/actions-runner under a dedicated user `runner`
(no sudo, not in build's group), rustup 1.99.0 pinned with both targets, sccache against /srv/sccache in READ_ONLY mode
on its own server port, Node 22 and mingw from the system, GitHub's svc.sh unit with a Nice 10 drop-in; registered on
igneum-network/igneum as igneum-build-1 (labels self-hosted, linux, x64, igneum-build-1) through
infra/build-server/runner/register.sh (gh as igneum-labs, the token on ssh stdin, never logged). Idempotent after
the env files moved behind svc.sh install (its env.sh rewrites them). libicu74 and python3-numpy added to APT.
- main's slots ruling: SLOTS default 2; remote-run.sh sets CARGO_BUILD_JOBS 90 when it holds the only taken slot and 45
when both are held, BR_MEASURE=1 takes the `measure` file exclusively and excludes builds (builds hold it shared),
lock files open in append mode (the old `exec {fd}>` truncated a busy slot's holder line on every probe), env
IGNEUM_BUILD_SLOTS_DIR and IGNEUM_BUILD_LOG_DIR win over the profile, `--self-test-slots` with five cases (the old
script fails it with JOBS=none); build-remote.sh and cross-remote.sh pass -j only when --jobs is given.
- infra/build-server/prover/cpu-trial.sh: the SP1 CPU prover on one fixture shard under the measure hold with a VmHWM
poller; 6 Oct 2026 run: core 34.2 s, compressed 85.9 s, peak RSS 28.2 GB on 96 threads, so no standing CPU prover.
- docs/plans/ci-self-hosted.md: the proposed runs-on change for ci.yml behind the repository variable IGNEUM_CI_RUNNER
(GitHub-hosted is the fallback), and why windows.yml cannot move to a Linux box. Workflows untouched.
- docs/plans/build-server.md section 7.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Rebased onto origin/master. Resolved: site/miner.html and site/miners.html taken whole from master (the site-miner copy,
the PC 1 images, the 42-character h1, lever 4 dated, the ember-tune "Measured by the team" row) and re-dressed in the shared
chrome by the build; site/build.mjs is master's (the shipper's downloads rule: the snapshot is written only on
SITE_DOWNLOADS_REFRESH=1, --refresh-downloads or CI; the priors team rows) plus the generated-page template on site.css and
the per-page eyebrow; site/index.html is the one-screen page with master's content ported in: the hero's proving sentence
(today's app on 24 GB, the 6 October rented-card measurement with its log link, "ships when the packaging row lands"), the
mine lead, the "Four pages" and "Ember Tune, measured" rows, the PC 1 figure and caption. The ledger generator's section
heading balances its lines and sits at 20 to 28 px so "Launch and operations" no longer leaves a word alone at 390 px.
Checks on this tree: identity grep 0 hits over 232 files, link check 884 links 0 broken, ledger text 43 of 43, contrast
32 pairs 0 under 4.5:1, orphan check 0 site headings (14 log titles reported), ledger page 0 leaks. Proof captures at 1440
dark: docs/plans/site-ui-3-shots/after/index-1440-dark.jpg and miner-1440-dark.jpg.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
site/site.css: the tokens of the miner app's third redesign in light and dark (a darker light-mode ember and molten for text
so every pair passes 4.5:1, checked by tools/ci/site-contrast-check.mjs), the type scale on Unbounded, Plex Sans and Plex
Mono, the nav with one primary, the footer with a System/Light/Dark control, cards, tiles, tables that scroll on a phone,
buttons, pills, notes, callouts, the state words, focus rings, reduced motion, print. Partials rebuilt; the head loads the
stylesheet and applies the stored theme before paint. Every page's own style block is reduced to its page rules.
site/live-dag.js: the shared DAG view for / and /live (docs/plans/site-ui-3.md section 6): blocks by header time and
miner lane, parent edges with the selected chain as one heavier path, blue lit, red dim, pending grey, proven filled,
checkpoint bands with their lock weight (locked solid, pending dashed), new blocks arriving with a glow, hover and tap
details in the ledger's words, wheel and pinch zoom of the window (30 to 300 s), a pause button, device-pixel sharpness,
a still frame under reduced motion; opts.mine marks the user's own blocks (the miner app embeds the same file),
opts.poll:false with dag.push(reply) lets a host feed it. Every pixel is a block the observer stored.
site/index.html: the one-screen story (what it is, the live scene with chain block, shards proven, last lock, vote keys and
hash rate as state words, three things a sceptic checks, the download, how it works, the wallet, the journey, economics),
every tick-list row kept, the chip model's numbers moved to the sceptic card, the testnet terms behind a chevron, the
scroll-reveal gone, the Open Graph set on the 1200 by 630 image, no horizontal overflow at 390.
site/litepaper.html: dark like the site with light on request, whole paper by default (the section mode kept, deep links
intact, print prints the paper), repository paths off the surface, the cover dated 6 October, the wallet at 0.1.4, the
roadmap's verifier figure aligned with the Mining section, the proof lag stated as measured (about 380 s against the 60 s
gate), the 0.3.6 plan dated with 0.3.14 stated, the two "tonight" placeholders said as not yet measured. Every ledger
sentence verbatim (tools/ci/ledger-text-check.mjs, 43 sentences).
site/live.html: the lane chart replaced by the module; "proven A/B in 10 min", "finality paused, last #N", "pending" for
"n/a"; the fee sentence true on both sides of DAA 210,000. site/metamask.html: its own canonical, the explorer linked, the
wallet's state said true. site/404.html: the page count and section count said true. site/build.mjs: the generated pages
on the system, each with its own eyebrow, the miners page's source notes as sentences. downloads.json refreshed from the
host (0.3.14). tools/ci/site-orphan-check.mjs: no site heading leaves one word alone at 390 px (log titles reported).
Checks: identity grep 0 hits, link check 854 links 0 broken, ledger text 43 of 43, contrast 32 pairs 0 under 4.5:1,
no horizontal overflow at 390 on 14 pages (the wallet page's timed table overflows by 5 px, for the wallet-ui-3 owner),
orphan check 0 site headings (the miner h1 at 96 characters is the site-miner agent's copy). After captures beside the
audit's: docs/plans/site-ui-3-shots/after/.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
docs/plans/site-ui-3-audit.md: the project lead's ask of 6 October 2026 ("run the website and all pages and litepaper through the same
apple lens as the miner app", "leave no stone unturned") against the miner-ui-3 standard. Method, screen by screen with the
10-point scale (home 6, litepaper 5, live 6, miner 5, wallet 6, miners 5, explorer 7, block and address 6, faucet 7,
metamask 6, 404 6, bench 5, evidence 6, ledger 5), links and downloads (914 attributes followed, four installers hashed
against the host), the live elements' loading, failed and stale states, contrast of every token pair in both themes,
keyboard focus, phone width, the litepaper's print, the top ten findings, and appendix B listing the stones turned.
docs/plans/site-ui-3-ticklist.md (appendix A): 502 rows across 17 inventories, every claim, number, date, link and
feature with its source and today's verdict (MATCH, STALE, UNSOURCED, APPROX), the ledger's stated sentences and their
presence, the thirteen "does not claim" sentences verbatim, the scrub rules.
docs/plans/site-ui-3.md: the design: one stylesheet (tokens light and dark, type on Unbounded, Plex Sans and Plex Mono,
the card and table rules, state words, nav, footer), the home page as one screen, the litepaper as a readable paper
with a sticky section nav and the whole paper by default, per-page notes, and the live DAG module brief (the project lead's "can this
look more advanced and cool?").
docs/plans/site-ui-3-shots/before/ and before-states/: every page at 1440 and 390, dark and light, the failed-fetch and
stale-observer captures, the litepaper print PDFs, the focus captures; headless Chromium (Playwright's build, never
Chrome.app), one browser at a time, niced.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Copy only, from origin/master c077caa (the fleet's record: docs/analysis/prover-tiers-real-cards.md and the bench-log
entry "Prover tiers on real cards"). The miner hero lead, the proving feat and the homepage lead each carry two
sentences: today's app with the stock SP1 server (a 24 GB NVIDIA card proves the full shard, RTX 4090 5.6 s and
RTX A5000 6.4 s; a 32 GB card mines and proves, RTX 5090 8.4 s; the stock server refuses 16 GB and under), and the
6 October 2026 measurement on eleven rented cards with the patched server (every NVIDIA card from 8 GB proves; 10 GB
and up mine and prove), linked to the bench-log entry and marked "ships when the packaging row lands".
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Audit fixes, copy only. (1) "every NVIDIA card from 8 GB proves" cited docs/analysis/prover-tiers-real-cards.md, which
is not in the repository, and contradicted evidence row 16; the hero lead, the proving feat, the three meta descriptions
and the homepage lead and h2 now say what row 16 and docs/analysis/amd-proving.md state: measured on an RTX 5090 with the
shipped SP1 server (13.9 GB floor), a 32 GB card mines and proves, a 24 GB card proves the full shard alone (from the
5090's allocation, not yet run on a 24 GB card), 16 GB empty shards only, 12 GB and under nothing on this build; the
eleven rented cards of 6 October are "measured, record pending". Each links /evidence#claims. (2) The h1 was 96
characters and seven lines on a phone; it is "Install. Start. The card mines and proves." (42) with the tiers in the lead.
(3) Lever row 4 said "Ships in 0.3.6" at 0.3.14; it now says shipped in 0.3.6 (the miner-latency merge, release-0.3.6.md
section 2), the app at 0.3.14.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The site agent committed the module at 3e5a880; the app's copy now matches it exactly (the earlier copy was taken
from the working tree before the commit). Same contract; quieter colours (chain in ember, included in bone, pending
in ash), opts.mine, opts.poll:false with dag.push.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
provision.sh step_cuda: NVIDIA's ubuntu2404 apt repository, cuda-nvrtc-dev-12-8, cuda-cudart-dev-12-8, cuda-driver-dev-12-8
(the libcuda stub) and opencl-c-headers; no nvcc (no build file calls it), no driver. tools/workers-remote.sh builds
igneum-worker-cuda and igneum-worker-opencl on the box from proto-cuda and proto-opencl with clang++ (static libstdc++),
prints sha256 and the glibc ceiling (2.38: fine for Ubuntu 24.04 hosts, not for 22.04 containers or HiveOS, where the Mac's
zig build stays). Proof: igneum-worker-cuda 1,613,992 B sha256 6db8a9ad..., igneum-worker-opencl 124,904 B sha256 0dea75bb...
remote-run.sh evaluates the command string in a subshell (a leaked set -e killed the runner after a successful build).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The public /ledger page (site/ledger.html) carried ~/.config/igneum paths five times, "pid 33114", --rpclisten=0.0.0.0:26610,
"PC 2" fifteen times, "the Mac" nineteen times and 202 repository file paths, because tools/ledger-page.mjs scrubbed with its
own short list and never ran the forbidden-strings hard stop (found by the site audit of 6 October 2026, docs/plans/site-ui-3-audit.md).
Now: the generator's scrub replaces every config or home-directory path with "a config file" or "a home-directory file", a pid with
"the process", a listen flag or 0.0.0.0 address with its plain words, "PC 1" and "PC 2" with "the Windows machine", "the Mac" with
"the Apple M5 Max" (the bench log's rule), and every repository file path with "a repository file"; the page's own source note
names no path. After rendering, the page is grepped with tools/ci/forbidden-strings.txt plus the leak classes and the render
exits 1 on a hit. The pattern list gains ~/.config, pid N, 0.0.0.0:port, PC 1 and PC 2 and --rpclisten=, and the identity grep
now covers site/ledger.html (html added to its file types). Regenerated: 167 entries, 0 leaks, identity grep 0 hits over 231
files, link check 0 broken.
Consequence per reader: the ledger keeps every criticism, status and answer; what a reader loses is the exact repository path
of a fix, which meant nothing outside the private repository. Nothing else on the site changes.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Main's ruling (6 October 2026, 20:1x UK): the box installs with the NUMBERS key alone so the 21:00 UK pulse comes from
it. install.sh accepts at least one key and names the missing ones; every tick's log line ends with "missing <keys>"; the
watcher without an incidents webhook logs its open or resolve and still advances its state, so the key landing later
does not flood the channel. Test added (31 passing).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
