Files
skills/public/hip-1145.html
zeekayandHanzo Dev 9e10c1c10e Each llms index points at the other two
hanzoskills.com/llms.txt now opens with the install line, the API
reference index and the company index, so an agent that lands on any
one of the three can find the rest.

robots.txt claimed 537 documents while llms.txt counts 920. The number
is gone from robots.txt; llms.txt derives it.

public/ is regenerated from scripts/site.py, which it had fallen
behind: the copy-prompt menu and agent bar were in the generator and
not in the committed pages. tests/site.mjs passes against the result.

Co-authored-by: Hanzo Dev <[email protected]>
2026-09-11 00:27:59 -07:00

259 lines
23 KiB
HTML

<!doctype html>
<html lang="en"><head>
<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">
<title>hip-1145</title>
<meta name="color-scheme" content="dark">
<style>
/* Committed to one visual world: a dark instrument panel. No light variant —
the dot field and the hairline rules only read on black, and a naive
inversion would wash both out. Every colour is painted explicitly so the page
never borrows a host background. */
@font-face{font-family:'Zen';src:url('/fonts/Zen-Variable.woff2') format('woff2');font-weight:100 900;font-style:normal;font-display:swap}
@font-face{font-family:'Zen';src:url('/fonts/Zen-Italic-Variable.woff2') format('woff2');font-weight:100 900;font-style:italic;font-display:swap}
@font-face{font-family:'Zen Mono';src:url('/fonts/ZenMono-Variable.woff2') format('woff2');font-weight:100 900;font-style:normal;font-display:swap}
@font-face{font-family:'Zen Mono';src:url('/fonts/ZenMono-Italic-Variable.woff2') format('woff2');font-weight:100 900;font-style:italic;font-display:swap}
@font-face{font-family:'Zen Pixel Square';src:url('/fonts/ZenPixel-Square.woff2') format('woff2');font-style:normal;font-display:swap}
:root{
--ink:#f4f4f2; --soft:#9c9c98; --faint:#6a6a66; --ground:#000; --panel:#0a0a0a;
--line:#232323; --line-hi:#3a3a3a; --accent:#f4f4f2;
--font-zen-sans:'Zen',-apple-system,BlinkMacSystemFont,"Segoe UI",Helvetica,Arial,sans-serif;
--font-zen-mono:'Zen Mono',ui-monospace,"SF Mono",SFMono-Regular,Menlo,Consolas,monospace;
--font-zen-pixel-square:'Zen Pixel Square',monospace;
--mono:var(--font-zen-mono);
--sans:var(--font-zen-sans);
/* Controls are pills and surfaces are soft, as on hanzo.ai. */
--r:999px; --r-soft:14px;
}
*{box-sizing:border-box}
html{-webkit-text-size-adjust:100%}
body{margin:0;background:var(--ground);color:var(--ink);font-family:var(--sans);
font-size:16px;line-height:1.65;-webkit-font-smoothing:antialiased}
.wrap{max-width:60rem;margin:0 auto;padding:0 1.75rem}
/* The dot field. A halftone made of one repeating radial-gradient, faded out
along the diagonal by a mask so it reads as a wave rather than wallpaper. */
/* No overflow clip here: the dot field is inset to the box already, and a clip
would cut off the copy-prompt menu. */
.hero{position:relative;border-bottom:1px solid var(--line)}
.hero::before{content:"";position:absolute;inset:0;
background-image:radial-gradient(var(--line-hi) 1px, transparent 1px);
background-size:7px 7px;
-webkit-mask-image:linear-gradient(115deg,#000 0%,rgba(0,0,0,.55) 38%,transparent 72%);
mask-image:linear-gradient(115deg,#000 0%,rgba(0,0,0,.55) 38%,transparent 72%);
pointer-events:none}
.hero .wrap{position:relative;padding-top:5.5rem;padding-bottom:4rem}
nav{border-bottom:1px solid var(--line)}
nav .in{max-width:60rem;margin:0 auto;padding:1.1rem 1.75rem;display:flex;
justify-content:space-between;align-items:center;gap:1rem}
.brand{font-family:var(--mono);font-size:.85rem;letter-spacing:.14em;
text-transform:uppercase;color:var(--ink);text-decoration:none;
display:inline-flex;align-items:center;gap:.6rem}
.brand i{font-style:normal;color:var(--faint)}
.mark{width:1.15em;height:1.15em;display:block;flex:0 0 auto}
nav a.up{font-family:var(--mono);font-size:.72rem;letter-spacing:.14em;
text-transform:uppercase;color:var(--soft);text-decoration:none}
nav a.up:hover{color:var(--ink)}
h1{font-family:var(--mono);font-size:clamp(2rem,5.4vw,3.4rem);line-height:1.12;
letter-spacing:-.02em;margin:0 0 1.1rem;font-weight:500;text-wrap:balance}
h1 em{font-style:normal;color:var(--faint)}
.lede{color:var(--soft);font-size:1.05rem;max-width:52ch;margin:0 0 2.25rem}
.btns{display:flex;flex-wrap:wrap;gap:.7rem;margin:0 0 .5rem;align-items:center}
.btn{display:inline-flex;align-items:center;gap:.5rem;font-family:var(--mono);
font-size:.72rem;letter-spacing:.14em;text-transform:uppercase;
padding:.8rem 1.4rem;border:1px solid var(--line-hi);border-radius:var(--r);
color:var(--ink);text-decoration:none;background:transparent;cursor:pointer;
line-height:1.2;font-weight:400}
.btn:hover{background:#141414;text-decoration:none;border-color:var(--soft)}
.btn.solid{background:var(--ink);color:#000;border-color:var(--ink)}
.btn.solid:hover{background:#d8d8d4}
.btn .chev{width:.7em;height:.7em;flex:0 0 auto;transition:transform .15s ease}
.btn[aria-expanded="true"] .chev{transform:rotate(180deg)}
/* Copy-prompt selector. Each row copies the prompt for its own target. */
.copy{position:relative;display:inline-flex}
.menu{position:absolute;top:calc(100% + .5rem);left:0;z-index:20;min-width:21rem;
background:var(--panel);border:1px solid var(--line-hi);border-radius:var(--r-soft);
padding:.4rem;display:flex;flex-direction:column;
box-shadow:0 12px 32px rgba(0,0,0,.75)}
.menu[hidden]{display:none}
.mi{display:flex;align-items:center;gap:.7rem;width:100%;text-align:left;
padding:.62rem .8rem;border:0;border-radius:calc(var(--r-soft) - .35rem);
background:transparent;color:var(--ink);font-family:var(--mono);font-size:.76rem;
letter-spacing:.04em;text-decoration:none;cursor:pointer;line-height:1.3;
white-space:nowrap}
.mi:hover,.mi:focus-visible{background:#181818;outline:none}
.mi .mark{width:1.05rem;height:1.05rem;color:var(--soft)}
.mi:hover .mark{color:var(--ink)}
.mi u{margin-left:auto;color:var(--faint);text-decoration:none;font-size:.68rem;
letter-spacing:.1em;text-transform:uppercase}
/* Works-with bar: one scrolling row, name and mark per agent. */
.strip{border-bottom:1px solid var(--line);background:var(--panel)}
.strip .in{max-width:60rem;margin:0 auto;padding:.9rem 1.75rem;display:flex;
align-items:center;gap:1.25rem;font-family:var(--mono);font-size:.72rem;
letter-spacing:.1em;text-transform:uppercase;color:var(--faint)}
.strip b{color:var(--soft);font-weight:400;flex:0 0 auto;
border-right:1px solid var(--line);padding-right:1.25rem;align-self:stretch;
display:flex;align-items:center}
.bar{display:flex;align-items:center;gap:.5rem;flex:1 1 auto;min-width:0;
overflow-x:auto;scrollbar-width:none;scroll-snap-type:x proximity;
-webkit-overflow-scrolling:touch;
-webkit-mask-image:linear-gradient(90deg,#000 calc(100% - 3rem),transparent);
mask-image:linear-gradient(90deg,#000 calc(100% - 3rem),transparent)}
.bar::-webkit-scrollbar{display:none}
.agent{display:inline-flex;align-items:center;gap:.45rem;flex:0 0 auto;
padding:.4rem .85rem;border:1px solid var(--line);border-radius:var(--r);
color:var(--soft);white-space:nowrap;scroll-snap-align:start;
text-decoration:none;background:var(--ground)}
.agent:hover{color:var(--ink);border-color:var(--line-hi);text-decoration:none}
.agent .mark{width:.95rem;height:.95rem}
h2{font-family:var(--mono);font-size:1.1rem;letter-spacing:-.01em;margin:2.75rem 0 .75rem;font-weight:500}
h3{font-size:1rem;margin:2rem 0 .5rem;font-weight:600}
h4,h5,h6{font-size:.94rem;margin:1.5rem 0 .5rem;font-weight:600}
p{margin:0 0 1rem;max-width:70ch}
ul,ol{margin:0 0 1rem;padding-left:1.35rem;max-width:70ch}
li{margin:.22rem 0}
a{color:var(--ink);text-decoration:underline;text-underline-offset:3px;
text-decoration-color:var(--line-hi)}
a:hover{text-decoration-color:var(--ink)}
a:focus-visible{outline:1px solid var(--ink);outline-offset:3px}
code{font-family:var(--mono);font-size:.85em;background:var(--panel);
border:1px solid var(--line);border-radius:5px;padding:.1em .36em}
pre{background:var(--panel);border:1px solid var(--line);padding:1rem 1.1rem;
border-radius:var(--r-soft);overflow-x:auto;margin:0 0 1.25rem}
pre code{background:none;border:0;padding:0;font-size:.82rem;line-height:1.7}
blockquote{margin:0 0 1rem;padding-left:1rem;border-left:1px solid var(--line-hi);color:var(--soft)}
hr{border:0;border-top:1px solid var(--line);margin:2.5rem 0}
table{border-collapse:collapse;width:100%;margin:0 0 1.25rem;font-size:.92rem;
display:block;overflow-x:auto}
th,td{border:1px solid var(--line);padding:.5rem .7rem;text-align:left}
.curl{display:block;background:var(--panel);border:1px solid var(--line);
border-radius:var(--r-soft);padding:.95rem 1.1rem;font-family:var(--mono);
font-size:.8rem;color:var(--soft);overflow-x:auto;margin:0 0 2.5rem;
white-space:pre;line-height:1.9}
.curl b{color:var(--ink);font-weight:400}
.curl i{font-style:normal;color:var(--faint)}
/* index */
#q{width:100%;padding:.85rem 1.2rem;font-family:var(--mono);font-size:.85rem;color:var(--ink);
background:var(--panel);border:1px solid var(--line);border-radius:var(--r);margin:0 0 .6rem}
#q::placeholder{color:var(--faint);text-transform:uppercase;letter-spacing:.1em;font-size:.72rem}
#q:focus{outline:none;border-color:var(--soft)}
#count{color:var(--faint);font-family:var(--mono);font-size:.72rem;letter-spacing:.1em;
text-transform:uppercase;margin:0 0 2rem}
.cat{margin:2.5rem 0 .5rem;font-family:var(--mono);font-size:.7rem;letter-spacing:.16em;
text-transform:uppercase;color:var(--faint);border-bottom:1px solid var(--line);
padding-bottom:.45rem}
.row{display:block;padding:.6rem .9rem;margin:0 -.9rem;border-bottom:1px solid var(--line);
border-radius:var(--r-soft);text-decoration:none}
.row:hover{background:#0a0a0a}
.nm{font-family:var(--mono);font-size:.84rem;color:var(--ink);display:block}
.ds{color:var(--soft);font-size:.9rem;display:block;margin-top:.12rem}
footer{margin-top:4rem;padding:2rem 0 4rem;border-top:1px solid var(--line);
color:var(--faint);font-size:.85rem}
footer a{color:var(--soft)}
.fmap{display:grid;gap:1.75rem;grid-template-columns:repeat(auto-fit,minmax(11rem,1fr));
margin:0 0 2rem;font-size:.85rem}
.fmap b{display:block;font-family:var(--mono);font-size:.68rem;letter-spacing:.16em;
text-transform:uppercase;color:var(--faint);font-weight:400;margin:0 0 .6rem;
border-bottom:1px solid var(--line);padding-bottom:.4rem}
.fmap a{display:inline-block;color:var(--ink);text-decoration:none;margin:.3rem .35rem 0 0}
.fmap a:hover{text-decoration:underline}
.fmap div{color:var(--faint);line-height:1.5}
.fend{border-top:1px solid var(--line);padding-top:1.25rem;font-size:.82rem}
@media (max-width:640px){
.hero .wrap{padding-top:3.25rem;padding-bottom:2.75rem}
.btns{gap:.5rem}
.btn{padding:.75rem 1.15rem;font-size:.68rem}
.copy,.copy .btn{width:100%;justify-content:center}
.menu{left:0;right:0;min-width:0}
.strip .in{flex-direction:column;align-items:flex-start;gap:.6rem;padding:.9rem 1.75rem}
.strip b{border-right:0;padding-right:0}
.bar{width:100%}
}
</style>
<link rel="alternate" type="text/markdown" href="/hip-1145.md">
<link rel="icon" href="/favicon.svg">
<link rel="canonical" href="https://hanzoskills.com/hip-1145">
</head>
<!-- Reading this with a machine? Every page here is also plain markdown.
This one: https://hanzoskills.com/hip-1145.md
Index: https://hanzoskills.com/llms.txt
Corpus: https://hanzoskills.com/all.md
Append .md to any path for the markdown twin. -->
<body>
<nav><div class="in">
<a class="brand" href="/"><svg class="mark" viewBox="0 0 67 67" fill="currentColor" aria-hidden="true"><path d="M22.21 67V44.6369H0V67H22.21Z"/><path d="M66.7038 22.3184H22.2534L0.0878906 44.6367H44.4634L66.7038 22.3184Z"/><path d="M22.21 0H0V22.3184H22.21V0Z"/><path d="M66.7198 0H44.5098V22.3184H66.7198V0Z"/><path d="M66.7198 67V44.6369H44.5098V67H66.7198Z"/></svg> HANZO<i>SKILLS</i></a>
<a class="up" href="/">All skills</a>
</div></nav>
<div class="wrap">
<h1>hip-1145</h1><p class="lede">HIP-1145: Research — The Experiment Record. Status Final. Hanzo architectural specification.</p><h1>HIP-1145: Research — The Experiment Record</h1>
<h2>Abstract</h2>
<p><code>/v1/research</code> is every experiment an org has ever run, kept and comparable: versioned, append-only evidence under one discriminator — <code>kind</code> ∈ benchmark | kernel-perf | training | ablation | policy-eval — with provenance (git sha, branch, dirty, library versions) as queryable columns. It is implemented in <code>hanzoai/cloud</code> at <code>apps/research</code>. A correction appends a new version under the same stable id and the prior version is RETAINED, never mutated; faulted and failed runs are retained too, because a negative result is evidence (<code>apps/research/research.go:17-22</code>).</p>
<h2>Motivation</h2>
<p>A measurement that cannot be re-found, re-attributed to its commit, or compared against its own history is an anecdote. The record exists so the question &quot;which library version regressed this benchmark&quot; is a query over the longitudinal record rather than an archaeology project — and so dedup never reads as loss: RETAINED is the full history, CANONICAL is the deterministic deduped view over it.</p>
<h2>Specification</h2>
<p>The key words MUST, MUST NOT and SHOULD are to be interpreted as in RFC 2119.</p>
<h3>Two planes, one owner</h3>
<p>Each org&#x27;s transactional SQLite — <code>cloud.OrgStore</code>, one physically separate <code>research.db</code> per org (<code>apps/research/research.go:191-206</code>) — is the local source of truth, riding whatever HA path the deployment configured. It rolls up best-effort into <code>hanzoai/datastore</code>, the platform&#x27;s column-oriented OLAP plane, for the cross-project query surface: losing a roll-up MUST never fail an ingest whose SQLite write already committed (<code>apps/research/datastore.go:11-16</code>). The stronger &quot;immutable · replicated · recovery-tested&quot; claim is deliberately NOT yet made — retry, reconciliation and backup/restore are still on the critical path (<code>apps/research/research.go:12-15</code>).</p>
<h3>Addresses</h3>
<p>Seventeen operations under <code>/v1/research</code>, sixteen typed and one declared (<code>apps/research/research.go</code>). The measurement plane: ingest (<code>POST /experiments</code>, idempotent by content — re-running a backfill appends nothing), the canonical listing (<code>GET /experiments</code>), <code>GET /projects</code>, <code>GET /totals</code>, the grant (<code>POST /grants</code>), and the diary (<code>POST /artifacts</code>, <code>GET /artifacts</code>, <code>GET /artifacts/{sha256}</code>). The last is the one route that cannot be a typed op: it streams the artifact&#x27;s raw bytes under the artifact&#x27;s own Content-Type, which a typed op — always JSON from a Go value — has no vocabulary for; its prose is declared beside the route.</p>
<p>The experiment record adds four kinds in one shape — a POST that records a batch, a GET that lists it narrowed by the filters that matter, with <code>?id=</code> selecting one so there is no second address for fetching one thing: <code>/benchmarks</code>, <code>/runs</code>, <code>/studies</code>, <code>/papers</code> — plus <code>GET /compare</code>.</p>
<p>An artifact&#x27;s identity is the SERVER-derived sha256 of its bytes, never a client-asserted hash, so the address space is genuinely content-addressed and un-poisonable; a re-POST of the same bytes is a no-op.</p>
<h3>The execution, and whether it finished</h3>
<p><code>Experiment</code> is one measured number and <code>Attempt</code> is one scored item; neither is the EXECUTION. <code>Run</code> is (<code>apps/research/record.go</code>): the benchmark and split it measured against, the system under test and its version, the configuration that moves a number (embedder, reader, k, temperature, max-tokens, the prompt file and its digest), the digests that pin its inputs (dataset, retrieval store, fact layer), the commit it was frozen at, its per-(category, metric) measures with the intervals the harness reported, when it ran, by whom — and <code>questions</code>, <code>answered</code> and <code>ended</code>.</p>
<p><code>completion</code> is DERIVED from those three and MUST NOT be accepted from a write: <code>complete</code> where the record says the run covered its set and says when it ended, <code>partial</code> where it says it covered less, <code>inconsistent</code> where the two counts cannot both be true, <code>unknown</code> where the record does not say. A harness that checkpoints as it goes reports whatever has been answered so far, so a run read at question 57 of 282 is otherwise indistinguishable from a finished one — which is how <code>locomo-all-context-k20-enso-flash</code> reached hanzo.ai/benchmarks 9.4 F1 points and 12.5 EM points above where it landed. Absence MUST stay absent: <code>questions</code>, <code>answered</code>, <code>k</code>, <code>temperature</code> and <code>max_tokens</code> are nullable, and a null is the record declining to answer rather than a zero.</p>
<p><code>ResearchBenchmark</code> is the TASK — dataset, that dataset&#x27;s licence and origin, the splits it declares, the metric its authors score it by, the citation — and it is independent of anyone&#x27;s run. <code>Study</code> is the question a set of runs was run to answer; <code>Paper</code> cites runs, and a run names at most its study, never a paper. A baseline is a run carrying <code>baseline</code>, not a kind of its own and not a category inside somebody else&#x27;s numbers.</p>
<p><code>GET /compare</code> reads two runs together and states FIRST what stops them being compared — a different benchmark, a different split, different dataset digests, either run not established as complete, either run not naming its system — then every configuration field they disagree about, then the change over what BOTH measured, with whether the reported intervals overlap. A metric only one of them reported is left out rather than compared against nothing.</p>
<p>All four ride the same per-org SQLite through the same orm records, the same content-hash version identity, the same server-assigned append clock and the same private-by-default grant: a correction to a run APPENDS a version and the published mistake stays retained. Runs also roll up to the warehouse with <code>completion</code> as a column; their per-category measures do not, because a second copy of the numbers with no reconciliation between them is worse than a join.</p>
<h3>Tenancy</h3>
<p>The org is the validated principal&#x27;s and selects the physical file (<code>storeFor</code> → <code>cloud.OrgStore</code>); the project is the server&#x27;s validated sub-scope, stamped positionally into every row and every warehouse write — a payload cannot forge either (<code>apps/research/datastore.go:31-32</code>). One org cannot read another&#x27;s evidence even in the presence of a query defect, because the other org&#x27;s rows are not in the file being queried.</p>
<h3>Consent is a separate grant</h3>
<p>An upload records <code>visibility=private</code> and grants NO training or commons-publication rights. Board visibility, <code>trainable</code> and <code>publishable</code> are each a SEPARATE authorized decision through <code>POST /v1/research/grants</code> — never implied by uploading a run (<code>apps/research/research.go:28-31</code>). Ingest MUST ignore any visibility or consent field a payload carries.</p>
<h3>The BYO endpoint is stored, not dialed</h3>
<p>An experiment may name the endpoint it was measured against. Ingest refuses the obviously-hostile URL (https only, every resolved address publicly routable — <code>apps/research/ssrf.go</code>), and that check is declared to be ingest hygiene, not the dial-time control: nothing in this capability dials a recorded endpoint, and code that ever does MUST carry a DialContext IP-pin first.</p>
<h3>Money, events, telemetry</h3>
<p>Free, said in those words: <code>plugin/research/main.go</code> declares <code>cloud.Free</code>. It publishes nothing to the bus, so a customer&#x27;s webhooks receive nothing from it. Beyond the request span, its only extra emission is the warehouse roll-up rows themselves — its own data plane, not telemetry a customer reads back under <code>/v1/o11y</code>.</p>
<h3>Stage</h3>
<p><code>beta</code>: the manifest row declares <code>Stage: Beta</code> (<code>manifest/apps.go:410</code>). The durability contract is still rolling out — the roll-up is best-effort and reconciliation is unbuilt, so the record is versioned-append-only today and no more is claimed.</p>
<h3>Loading a measurement tree</h3>
<p><code>research.Load</code> reads a tree&#x27;s OWN output — the harness&#x27;s generated <code>benchmarks.json</code> for the numbers and each run directory&#x27;s <code>meta.json</code> and <code>metrics.json</code> for what the run was — and recomputes no metric: a second normalization would be a second answer to the same question, free to disagree with the one the published tables came from. Where a field is absent it is recorded absent and counted, never defaulted. Measured over <code>bench/brain</code> as it stands: 56 runs and 2936 measures, of which 9 are complete, 3 are inconsistent (their own <code>meta.json</code> records more answers than questions), and 44 are unknown; no run records who ran it or the version of what it measured; 44 carry no completion counts, 28 none of the reader knobs, 4 no commit.</p>
<h3>Upstream</h3>
<p>Derives from none. The OLAP half reuses the platform&#x27;s own <code>hanzoai/datastore</code> connection beside the account-usage warehouse; nothing external is forked, embedded or mirrored.</p>
<h2>Rationale</h2>
<p>Versioned-retained, rather than mutate-in-place with an audit log, is the load bearing choice: an evidence plane that can silently replace a number is a marketing plane (HIP-0512 makes the same argument for verdicts). Idempotency by content rather than by a client key follows — a backfill has no stable client key, and content identity makes replays free. Consent as a separate grant, rather than a flag on upload, exists because the uploader of a run is routinely a CI job with no authority to publish or license anything.</p>
<h2>Security Considerations</h2>
<p>The record carries two things worth stealing and one thing worth forging. Stealing: another org&#x27;s benchmark evidence (closed physically, per-org files) and an artifact&#x27;s raw bytes, which may carry licensed or personal material — the artifact read is org-scoped before the hash is looked up, so a known hash alone retrieves nothing across tenants. Forging: consent. The wrong implementation lets an upload set <code>trainable:true</code> or a client assert an artifact&#x27;s hash — the first turns a CI credential into a licensing authority, the second poisons the content address so a later honest upload resolves to attacker bytes. Ingest forcing private/withheld and the server deriving every hash are the closures. The stored BYO endpoint is the residual SSRF surface, held closed by nothing dialing it.</p>
<h2>References</h2>
<ul><li>HIP-0026 — Identity and Access Management</li><li>HIP-0106 — The Hanzo Plugin Contract</li><li>HIP-0139 — Capability</li><li>HIP-0512 — Experiment — The Evidence Plane</li></ul>
<h2>Copyright</h2>
<p>Released under CC0 1.0 Universal Public Domain Dedication.</p>
<footer>
<div class="fmap">
<div><b>Build</b>
<a href="https://hanzo.app">hanzo.app</a> app builder
<a href="https://hanzo.codes">hanzo.codes</a> code with an agent
<a href="https://hanzoskills.com">hanzoskills.com</a> what an agent reads</div>
<div><b>Run</b>
<a href="https://console.hanzo.ai">console.hanzo.ai</a> cloud console
<a href="https://api.hanzo.ai">api.hanzo.ai</a> the API</div>
<div><b>Account</b>
<a href="https://hanzo.id">hanzo.id</a> sign in
<a href="https://pay.hanzo.ai">pay.hanzo.ai</a> top up
<a href="https://billing.hanzo.ai">billing.hanzo.ai</a> invoices and usage</div>
<div><b>Company</b>
<a href="https://hanzo.ai">hanzo.ai</a> canonical
<a href="https://hanzo.ventures">hanzo.ventures</a> ventures</div>
</div>
<div class="fend"><a href="/hip-1145.md">This page as plain text</a> · <a href="/llms.txt">Index for agents</a></div>
</footer>
</div></body></html>