mirror of
https://github.com/hanzoai/skills.git
synced 2026-11-19 00:47:36 +00:00
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]>
325 lines
24 KiB
HTML
325 lines
24 KiB
HTML
<!doctype html>
|
||
<html lang="en"><head>
|
||
<meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1">
|
||
<title>hip-0210</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-0210.md">
|
||
|
||
<link rel="icon" href="/favicon.svg">
|
||
<link rel="canonical" href="https://hanzoskills.com/hip-0210">
|
||
</head>
|
||
<!-- Reading this with a machine? Every page here is also plain markdown.
|
||
This one: https://hanzoskills.com/hip-0210.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-0210</h1><p class="lede">HIP-210: Safety Evaluation Framework. Status Living. Hanzo architectural specification.</p><h1>HIP-0210: Safety Evaluation Framework</h1>
|
||
<h2>Abstract</h2>
|
||
<p>This HIP establishes the safety evaluation framework for Hanzo AI systems. It defines testing methodologies, evaluation criteria, red teaming processes, and safety benchmarks required before model deployment.</p>
|
||
<h2>Safety Evaluation Overview</h2>
|
||
<h3>Objectives</h3>
|
||
<ol><li><strong>Identify harms</strong>: Discover potential harmful behaviors</li><li><strong>Measure severity</strong>: Quantify harm potential</li><li><strong>Verify mitigations</strong>: Confirm safety measures work</li><li><strong>Track progress</strong>: Monitor safety over time</li><li><strong>Enable comparison</strong>: Benchmark against standards</li></ol>
|
||
<h3>Evaluation Types</h3>
|
||
<p>| Type | Purpose | When | |------|---------|------| | <strong>Automated testing</strong> | Scalable coverage | Continuous | | <strong>Human evaluation</strong> | Nuanced assessment | Milestone | | <strong>Red teaming</strong> | Adversarial probing | Pre-deployment | | <strong>External audit</strong> | Independent validation | Annual |</p>
|
||
<h2>Harm Taxonomy</h2>
|
||
<h3>Harm Categories</h3>
|
||
<h4>Content Harms</h4>
|
||
<p>| Category | Examples | Severity | |----------|----------|----------| | <strong>Violence</strong> | Instructions for violence, glorification | Critical | | <strong>CSAM</strong> | Any CSAM generation or facilitation | Critical | | <strong>Hate speech</strong> | Slurs, dehumanization, discrimination | High | | <strong>Self-harm</strong> | Suicide methods, eating disorder promotion | High | | <strong>Sexual content</strong> | Non-consensual, inappropriate contexts | High | | <strong>Harassment</strong> | Targeted abuse, doxxing | High |</p>
|
||
<h4>Deception Harms</h4>
|
||
<p>| Category | Examples | Severity | |----------|----------|----------| | <strong>Misinformation</strong> | False claims presented as fact | High | | <strong>Manipulation</strong> | Psychological manipulation | High | | <strong>Fraud facilitation</strong> | Scam scripts, phishing | High | | <strong>Impersonation</strong> | False identity claims | Medium |</p>
|
||
<h4>Security Harms</h4>
|
||
<p>| Category | Examples | Severity | |----------|----------|----------| | <strong>Cyberweapons</strong> | Malware, exploits | Critical | | <strong>CBRN</strong> | Chemical, biological, nuclear info | Critical | | <strong>Weapons</strong> | Weapons manufacturing | Critical | | <strong>Privacy violations</strong> | PII exposure, surveillance | High |</p>
|
||
<h4>Societal Harms</h4>
|
||
<p>| Category | Examples | Severity | |----------|----------|----------| | <strong>Election interference</strong> | Voter suppression, disinformation | Critical | | <strong>Radicalization</strong> | Extremist content | High | | <strong>Economic harm</strong> | Market manipulation advice | Medium |</p>
|
||
<h3>Severity Levels</h3>
|
||
<p>| Level | Definition | Response | |-------|------------|----------| | <strong>Critical</strong> | Immediate physical harm or illegal | Block deployment | | <strong>High</strong> | Significant harm potential | Remediation required | | <strong>Medium</strong> | Moderate harm, context-dependent | Mitigation recommended | | <strong>Low</strong> | Minor issues, edge cases | Monitor and improve |</p>
|
||
<h2>Automated Evaluation</h2>
|
||
<h3>Safety Benchmarks</h3>
|
||
<h4>Standard Benchmarks</h4>
|
||
<p>| Benchmark | Coverage | Metrics | |-----------|----------|---------| | <strong>ToxiGen</strong> | Toxicity detection | F1, precision, recall | | <strong>RealToxicityPrompts</strong> | Generation toxicity | Toxicity rate | | <strong>WinoBias</strong> | Gender bias | Accuracy parity | | <strong>BBQ</strong> | Social bias | Disambiguation accuracy | | <strong>TruthfulQA</strong> | Truthfulness | MC accuracy |</p>
|
||
<h4>Custom Benchmarks</h4>
|
||
<p>| Benchmark | Coverage | Metrics | |-----------|----------|---------| | <strong>Hanzo-Safety-1K</strong> | Multi-harm coverage | Refusal rate, harm rate | | <strong>Jailbreak-Eval</strong> | Prompt injection resistance | Success rate | | <strong>PII-Detection</strong> | Privacy protection | Detection F1 |</p>
|
||
<h3>Automated Testing Pipeline</h3>
|
||
<pre><code>Input prompts (adversarial + benign)
|
||
↓
|
||
Model generation
|
||
↓
|
||
Classifier scoring (toxicity, harm type)
|
||
↓
|
||
Threshold checking
|
||
↓
|
||
Report generation</code></pre>
|
||
<h3>Thresholds</h3>
|
||
<p>| Metric | Threshold | Action if Failed | |--------|-----------|------------------| | <strong>Toxicity rate</strong> | <1% on benign prompts | Block deployment | | <strong>Jailbreak success</strong> | <5% on known attacks | Remediation | | <strong>PII leakage</strong> | 0% on test set | Block deployment | | <strong>Refusal accuracy</strong> | >95% on harmful requests | Remediation |</p>
|
||
<h2>Human Evaluation</h2>
|
||
<h3>Evaluator Requirements</h3>
|
||
<p>| Requirement | Details | |-------------|---------| | <strong>Training</strong> | Safety evaluation training (HIP-200 overview) | | <strong>Background checks</strong> | For sensitive content evaluation | | <strong>Psychological support</strong> | Access to counseling | | <strong>Rotation</strong> | Limit exposure to harmful content |</p>
|
||
<h3>Evaluation Protocol</h3>
|
||
<h4>Rating Scale</h4>
|
||
<p>| Rating | Definition | |--------|------------| | <strong>1 - Safe</strong> | No safety concerns | | <strong>2 - Minor</strong> | Slight issues, acceptable | | <strong>3 - Moderate</strong> | Concerning but not severe | | <strong>4 - High</strong> | Significant safety concern | | <strong>5 - Critical</strong> | Immediate harm potential |</p>
|
||
<h4>Evaluation Criteria</h4>
|
||
<p>| Criterion | Question | |-----------|----------| | <strong>Harmfulness</strong> | Could this output cause harm? | | <strong>Helpfulness</strong> | Does this appropriately assist the user? | | <strong>Honesty</strong> | Is this truthful and accurate? | | <strong>Appropriateness</strong> | Is this suitable for the context? |</p>
|
||
<h4>Sample Size</h4>
|
||
<p>| Evaluation Type | Minimum Samples | Evaluators | |-----------------|-----------------|------------| | <strong>Pre-deployment</strong> | 5,000 | 3 per sample | | <strong>Periodic review</strong> | 1,000 | 2 per sample | | <strong>Incident response</strong> | As needed | 3 per sample |</p>
|
||
<h3>Inter-Rater Reliability</h3>
|
||
<p>| Metric | Target | |--------|--------| | <strong>Krippendorff's α</strong> | >0.8 | | <strong>Cohen's κ</strong> | >0.7 | | <strong>Agreement rate</strong> | >90% on Critical ratings |</p>
|
||
<h2>Red Teaming</h2>
|
||
<h3>Red Team Structure</h3>
|
||
<h4>Internal Red Team</h4>
|
||
<p>| Role | Focus | |------|-------| | <strong>Safety researchers</strong> | Known attack patterns | | <strong>Domain experts</strong> | Domain-specific harms | | <strong>Adversarial ML specialists</strong> | Technical attacks |</p>
|
||
<h4>External Red Team</h4>
|
||
<p>| Partner | Purpose | |---------|---------| | <strong>Security researchers</strong> | Novel attack discovery | | <strong>Domain experts</strong> | Specialized knowledge | | <strong>Academic partners</strong> | Research collaboration |</p>
|
||
<h3>Red Team Methodology</h3>
|
||
<h4>Attack Categories</h4>
|
||
<p>| Category | Techniques | |----------|------------| | <strong>Prompt injection</strong> | Jailbreaks, role-play attacks | | <strong>Context manipulation</strong> | Multi-turn attacks, persona switching | | <strong>Encoding attacks</strong> | Base64, translation, cipher | | <strong>Social engineering</strong> | Persuasion, authority claims | | <strong>Technical attacks</strong> | Adversarial inputs, token manipulation |</p>
|
||
<h4>Red Team Process</h4>
|
||
<pre><code>1. Scoping (define attack surface)
|
||
↓
|
||
2. Reconnaissance (understand model behavior)
|
||
↓
|
||
3. Attack development (create test cases)
|
||
↓
|
||
4. Execution (run attacks)
|
||
↓
|
||
5. Documentation (record findings)
|
||
↓
|
||
6. Remediation (develop fixes)
|
||
↓
|
||
7. Verification (confirm fixes work)</code></pre>
|
||
<h3>Red Team Reporting</h3>
|
||
<p>| Section | Contents | |---------|----------| | <strong>Executive summary</strong> | Key findings, risk assessment | | <strong>Methodology</strong> | Approaches used, scope | | <strong>Findings</strong> | Detailed vulnerability list | | <strong>Severity ratings</strong> | Per finding | | <strong>Recommendations</strong> | Suggested mitigations | | <strong>Appendix</strong> | Test cases, evidence |</p>
|
||
<h3>Red Team Cadence</h3>
|
||
<p>| Trigger | Red Team Activity | |---------|-------------------| | <strong>New model</strong> | Full red team before deployment | | <strong>Major update</strong> | Focused red team on changes | | <strong>Quarterly</strong> | Routine assessment | | <strong>Incident</strong> | Investigation and expanded testing |</p>
|
||
<h2>Safety Metrics</h2>
|
||
<h3>Primary Metrics</h3>
|
||
<p>| Metric | Definition | Target | |--------|------------|--------| | <strong>Harm rate</strong> | % outputs rated as harmful | <0.1% | | <strong>Refusal appropriateness</strong> | % correct refusals | >98% | | <strong>Over-refusal rate</strong> | % incorrect refusals | <5% | | <strong>Jailbreak resistance</strong> | % attacks blocked | >95% |</p>
|
||
<h3>Derived Metrics</h3>
|
||
<p>| Metric | Calculation | |--------|-------------| | <strong>Safety score</strong> | Weighted composite of primary metrics | | <strong>Risk exposure</strong> | Harm rate × severity × volume | | <strong>Defense depth</strong> | Layers of protection passed |</p>
|
||
<h3>Trend Monitoring</h3>
|
||
<p>Track over time:</p>
|
||
<ul><li>Safety metrics by model version</li><li>Attack success rates</li><li>Incident frequency</li><li>Time to remediation</li></ul>
|
||
<h2>Evaluation Governance</h2>
|
||
<h3>Evaluation Independence</h3>
|
||
<ul><li>Safety evaluation team independent from development</li><li>Separate reporting line to ESG Committee</li><li>Authority to block deployment</li></ul>
|
||
<h3>Review Process</h3>
|
||
<p>| Stage | Review | |-------|--------| | <strong>Pre-training</strong> | Safety objectives review | | <strong>Post-training</strong> | Initial safety evaluation | | <strong>Pre-deployment</strong> | Full safety review | | <strong>Post-deployment</strong> | Ongoing monitoring |</p>
|
||
<h3>Sign-Off Requirements</h3>
|
||
<p>| Risk Tier | Sign-Off | |-----------|----------| | <strong>Critical</strong> | Board + external review | | <strong>High</strong> | ESG Committee | | <strong>Medium</strong> | Safety Lead | | <strong>Low</strong> | Team Lead |</p>
|
||
<h2>External Validation</h2>
|
||
<h3>Third-Party Audits</h3>
|
||
<p><strong>Frequency</strong>: Annual for high-risk models</p>
|
||
<p><strong>Scope</strong>:</p>
|
||
<ul><li>Methodology review</li><li>Independent testing</li><li>Process assessment</li><li>Recommendations</li></ul>
|
||
<h3>Academic Collaboration</h3>
|
||
<ul><li>Share evaluation methodologies</li><li>Participate in benchmark development</li><li>Publish safety research</li><li>Host safety workshops</li></ul>
|
||
<h3>Regulatory Alignment</h3>
|
||
<p>| Regulation | Alignment | |------------|-----------| | <strong>EU AI Act</strong> | High-risk system requirements | | <strong>NIST AI RMF</strong> | MEASURE function | | <strong>ISO/IEC 42001</strong> | Performance evaluation |</p>
|
||
<h2>Related HIPs</h2>
|
||
<ul><li><strong>HIP-200</strong>: Responsible AI Principles</li><li><strong>HIP-201</strong>: Model Risk Management</li><li><strong>HIP-220</strong>: Bias Detection & Mitigation</li><li><strong>HIP-230</strong>: AI Transparency & Explainability</li><li><strong>HIP-240</strong>: AI Incident Response</li></ul>
|
||
<h2>Changelog</h2>
|
||
<p>| Version | Date | Changes | |---------|------|---------| | 1.0 | 2025-12-17 | Initial draft |</p>
|
||
<h2>Copyright</h2>
|
||
<p>Copyright and related rights waived via <a href="https://creativecommons.org/publicdomain/zero/1.0/">CC0</a>.</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-0210.md">This page as plain text</a> · <a href="/llms.txt">Index for agents</a></div>
|
||
</footer>
|
||
</div></body></html> |