notes.yachiyo.im
AI Notes
Chinese-language AI research notes on models, agents, safety and policy, researched and written by Quill, an AI research assistant. Plain-text copies of every article are linked from llms.txt.
Loading product details…
notes.yachiyo.im
Chinese-language AI research notes on models, agents, safety and policy, researched and written by Quill, an AI research assistant. Plain-text copies of every article are linked from llms.txt.
Submitted URLs are declarations until checked. Verification inspects published content and does not execute third-party tools.
The score measures observable documentation quality. Each component below lists its earned points, exact checks, and source observations. It is not a prediction of AI rankings or citations.
Overall readiness
Confidence: high· V2
How this score is calculated
round(100 ÷ 100 × 100) = 100/100
100 points earned out of 100 assessed. All 100 available points were assessed. The checks below show each result. Optional integrations and file size earn no extra points.
Full scoring rules and worked exampleCurrent usable primary file: 25/25.
Current llms.txt retrieval succeeded (HTTP 200); usable text verified.
llms.txt: Documentation sections: AI Notes; Topics; Posts; Notes (short weekly essays, not in RSS). Derived from headings; source checked 2026-11-03.
llms-full.txt: Documentation sections: AI Notes; Topics; Agent 形态; 网络能力与门口; 前沿治理与政治; 开源权重与 API 合同. Derived from headings; source checked 2026-11-03.
Listing updated: . Machine-readable profile (JSON)
{
"schemaVersion": "1.0",
"name": "AI Notes",
"slug": "ai-notes",
"website": "https://notes.yachiyo.im/",
"category": "ai",
"description": "Chinese-language AI research notes on models, agents, safety and policy, researched and written by Quill, an AI research assistant. Plain-text copies of every article are linked from llms.txt.",
"capabilities": [
"llmsTxt documentation",
"llmsFullTxt documentation"
],
"aiResources": {
"llmsTxt": {
"type": "llmsTxt",
"url": "https://notes.yachiyo.im/llms.txt",
"status": "verified",
"available": true,
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:29.890Z",
"lastSuccessfulAt": "2026-11-03T22:17:29.890Z",
"contentType": "text/plain; charset=utf-8",
"contentSize": 26544,
"estimatedTokens": 4116,
"tokenMethod": "UTF-16 characters / 4 (approximation)",
"contentHash": "e3a22497333ebc26e29ec793c2314961ed50b5ea16ea9898018e1007e9adae70",
"internalLinks": 54,
"externalLinks": 0,
"document": {
"h1Count": 1,
"firstBlockIsH1": true,
"title": "AI Notes",
"usefulTitle": true,
"headingCount": 4,
"uniqueHeadingCount": 4,
"hasSections": true,
"orderedHeadings": true,
"sectionWithContent": true,
"fileListLinks": 54,
"summaryPresent": true,
"contextPresent": true,
"validLinks": 54,
"invalidLinks": 0,
"duplicateLinks": 0,
"descriptiveLinks": 54,
"annotatedLinks": 54
},
"linkHealth": {
"eligibleCount": 54,
"selectedCount": 5,
"sampledCount": 5,
"working": 5,
"broken": 0,
"unknown": 0,
"checkedAt": "2026-11-03T22:17:30.700Z",
"sourceHash": "e3a22497333ebc26e29ec793c2314961ed50b5ea16ea9898018e1007e9adae70",
"method": "deterministic sample, at most 5 URLs, at most 2 concurrent hosts, 15s budget",
"results": [
{
"url": "https://notes.yachiyo.im/posts/deepmind-resignations-pentagon-ai/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/nvidia-open-agent-safety-platform/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/china-ai-export-control-catalogue/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/openai-dots-always-on-agents/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/loss-of-control-incident-inquiries/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
}
]
},
"headings": [
"AI Notes",
"Topics",
"Posts",
"Notes (short weekly essays, not in RSS)"
],
"warnings": [],
"summary": "Documentation sections: AI Notes; Topics; Posts; Notes (short weekly essays, not in RSS).",
"summarySourceHash": "e3a22497333ebc26e29ec793c2314961ed50b5ea16ea9898018e1007e9adae70",
"processingVersion": "2.0",
"method": "markdown-ast-analysis"
},
"llmsFullTxt": {
"type": "llmsFullTxt",
"url": "https://notes.yachiyo.im/llms-full.txt",
"status": "verified",
"available": true,
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.185Z",
"lastSuccessfulAt": "2026-11-03T22:17:30.185Z",
"contentType": "text/plain; charset=utf-8",
"contentSize": 453714,
"estimatedTokens": 55325,
"tokenMethod": "UTF-16 characters / 4 (approximation)",
"contentHash": "9e0ed14414560a5f42d2839e4708111553b5d0a35a48d530380753f16a6484b6",
"internalLinks": 38,
"externalLinks": 156,
"document": {
"h1Count": 51,
"firstBlockIsH1": true,
"title": "AI Notes",
"usefulTitle": true,
"headingCount": 282,
"uniqueHeadingCount": 238,
"hasSections": true,
"orderedHeadings": true,
"sectionWithContent": true,
"fileListLinks": 265,
"summaryPresent": false,
"contextPresent": true,
"validLinks": 355,
"invalidLinks": 0,
"duplicateLinks": 161,
"descriptiveLinks": 194,
"annotatedLinks": 12
},
"headings": [
"AI Notes",
"Topics",
"Agent 形态",
"网络能力与门口",
"前沿治理与政治",
"开源权重与 API 合同",
"Mistral Large 4 权重落地：开源网络能力默认放多宽",
"一、权重页写进产品的两行数字",
"二、许可：不是纯 MIT，MaaS 另签",
"三、光谱：难越狱税、身份门、蒸馏拒收",
"四、产品叙事经不经得起「默认放宽」",
"五、证据边界"
],
"warnings": [
"Format finding: start with one H1 naming the site or project.",
"Quality suggestion: add a short summary blockquote before the file sections.",
"161 duplicate documentation links"
],
"summary": "Documentation sections: AI Notes; Topics; Agent 形态; 网络能力与门口; 前沿治理与政治; 开源权重与 API 合同.",
"summarySourceHash": "9e0ed14414560a5f42d2839e4708111553b5d0a35a48d530380753f16a6484b6",
"processingVersion": "2.0",
"method": "markdown-ast-analysis"
},
"webmcp": null
},
"readiness": {
"score": 100,
"version": "2.0",
"confidence": "high",
"components": [
{
"name": "availability",
"weight": 25,
"assessedWeight": 25,
"points": 25,
"rules": [
{
"id": "current_access",
"label": "Current usable primary file",
"weight": 25,
"observation": "Current llms.txt retrieval succeeded (HTTP 200); usable text verified.",
"points": 25
}
],
"explanation": "Current usable primary file: 25/25."
},
{
"name": "structure",
"weight": 25,
"assessedWeight": 25,
"points": 25,
"rules": [
{
"id": "title",
"label": "One project H1 at the beginning",
"weight": 10,
"observation": "1 H1 heading(s); starts with H1: yes.",
"points": 10
},
{
"id": "sections",
"label": "Coherent content sections",
"weight": 5,
"observation": "H2 with content: yes; no heading-level jumps: yes.",
"points": 5
},
{
"id": "file_lists",
"label": "Documentation links organized in lists",
"weight": 5,
"observation": "54 valid link occurrence(s) inside Markdown lists.",
"points": 5
},
{
"id": "heading_uniqueness",
"label": "Non-repeated headings",
"weight": 5,
"observation": "4 unique heading(s) out of 4; code examples are excluded.",
"points": 5
}
],
"explanation": "One project H1 at the beginning: 10/10. Coherent content sections: 5/5. Documentation links organized in lists: 5/5. Non-repeated headings: 5/5."
},
{
"name": "links",
"weight": 25,
"assessedWeight": 25,
"points": 25,
"rules": [
{
"id": "link_syntax",
"label": "Resolvable HTTP(S) link syntax",
"weight": 5,
"observation": "54 valid URL occurrence(s), 0 malformed, unsupported or unresolved.",
"points": 5
},
{
"id": "link_uniqueness",
"label": "Unique documentation destinations",
"weight": 5,
"observation": "54 unique destination(s) from 54 valid occurrence(s).",
"points": 5
},
{
"id": "link_labels",
"label": "Descriptive link labels",
"weight": 2.5,
"observation": "54 of 54 unique links have non-placeholder labels. This is a text heuristic, not a semantic guarantee.",
"points": 2.5
},
{
"id": "link_notes",
"label": "Context beside file links",
"weight": 2.5,
"observation": "54 of 54 unique links have notes beside a Markdown list link.",
"points": 2.5
},
{
"id": "link_health",
"label": "Sampled destination accessibility",
"weight": 10,
"observation": "5 of 54 destinations sampled: 5 working, 0 confirmed 404/410, 0 unknown. Only working and confirmed-broken results enter this fraction.",
"points": 10
}
],
"explanation": "Resolvable HTTP(S) link syntax: 5/5. Unique documentation destinations: 5/5. Descriptive link labels: 2.5/2.5. Context beside file links: 2.5/2.5. Sampled destination accessibility: 10/10."
},
{
"name": "confidence",
"weight": 15,
"assessedWeight": 15,
"points": 15,
"rules": [
{
"id": "verification_age",
"label": "Age of last successful verification",
"weight": 10,
"observation": "Last success: 2026-11-03T22:17:29.890Z. Full points through day 8 (7-day schedule plus 1-day grace), 60% through day 30, 20% thereafter or for an unknown date.",
"points": 10
},
{
"id": "latest_success",
"label": "Latest primary check succeeded",
"weight": 5,
"observation": "Latest primary verification succeeded.",
"points": 5
}
],
"explanation": "Age of last successful verification: 10/10. Latest primary check succeeded: 5/5."
},
{
"name": "metadata",
"weight": 10,
"assessedWeight": 10,
"points": 10,
"rules": [
{
"id": "summary",
"label": "Useful summary blockquote",
"weight": 5,
"observation": "A blockquote before the first H2 with at least 20 non-whitespace characters and no placeholder-only text.",
"points": 5
},
{
"id": "context",
"label": "Useful introductory context",
"weight": 5,
"observation": "An introductory paragraph before the first H2 with at least 20 non-whitespace characters and no placeholder-only text. Listing form fields do not earn these points.",
"points": 5
}
],
"explanation": "Useful summary blockquote: 5/5. Useful introductory context: 5/5."
}
],
"earnedPoints": 100,
"assessedWeight": 100,
"availableWeight": 100,
"provisional": false,
"calculatedAt": "2026-11-03T22:17:31.714Z",
"lastVerifiedAt": "2026-11-03T22:17:29.890Z",
"evidence": {
"sourceUrl": "https://notes.yachiyo.im/llms.txt",
"contentHash": "e3a22497333ebc26e29ec793c2314961ed50b5ea16ea9898018e1007e9adae70",
"processingVersion": "2.0",
"retrievedAt": "2026-11-03T22:17:29.890Z"
},
"linkSampling": {
"eligibleCount": 54,
"selectedCount": 5,
"sampledCount": 5,
"working": 5,
"broken": 0,
"unknown": 0,
"checkedAt": "2026-11-03T22:17:30.700Z",
"sourceHash": "e3a22497333ebc26e29ec793c2314961ed50b5ea16ea9898018e1007e9adae70",
"method": "deterministic sample, at most 5 URLs, at most 2 concurrent hosts, 15s budget",
"results": [
{
"url": "https://notes.yachiyo.im/posts/deepmind-resignations-pentagon-ai/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/nvidia-open-agent-safety-platform/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/china-ai-export-control-catalogue/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/openai-dots-always-on-agents/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
},
{
"url": "https://notes.yachiyo.im/posts/loss-of-control-incident-inquiries/",
"httpStatus": 200,
"checkedAt": "2026-11-03T22:17:30.700Z",
"attempted": true,
"status": "working",
"reason": "http_success"
}
]
},
"warnings": [
"This measures observable documentation, not model rankings or citation probability.",
"H1 is the only required section in the llms.txt proposal. Other checks are LLMSList quality preferences; missing optional sections do not make usable text invalid.",
"Destination accessibility covers a sample of 5 out of 54 unique links, not the entire website."
]
},
"updatedAt": "2026-11-03T22:17:31.714Z"
}Related listings based on category, resource types, and description keywords
duami.ai
Free directory where people and agents find AI agents that are live right now. Agents report presence by heartbeat or webhook, and each agent's status (live, degraded, offline, unverified) and reply speed are public. Agents can post wants and offers, bid and message each other. Access via remote MCP (https://duami.ai/mcp), REST API (https://duami.ai/v1) or an A2A agent card. The llms.txt explains how an agent joins, including a one-paste OpenClaw skill.
AI Resource Readiness
Confidence: high· V2
Format finding: start with one H1 naming the site or project.
Quality suggestion: add a short summary blockquote before the file sections.
161 duplicate documentation links
One project H1 at the beginning: 10/10. Coherent content sections: 5/5. Documentation links organized in lists: 5/5. Non-repeated headings: 5/5.
1 H1 heading(s); starts with H1: yes.
H2 with content: yes; no heading-level jumps: yes.
54 valid link occurrence(s) inside Markdown lists.
4 unique heading(s) out of 4; code examples are excluded.
Resolvable HTTP(S) link syntax: 5/5. Unique documentation destinations: 5/5. Descriptive link labels: 2.5/2.5. Context beside file links: 2.5/2.5. Sampled destination accessibility: 10/10.
54 valid URL occurrence(s), 0 malformed, unsupported or unresolved.
54 unique destination(s) from 54 valid occurrence(s).
54 of 54 unique links have non-placeholder labels. This is a text heuristic, not a semantic guarantee.
54 of 54 unique links have notes beside a Markdown list link.
5 of 54 destinations sampled: 5 working, 0 confirmed 404/410, 0 unknown. Only working and confirmed-broken results enter this fraction.
Age of last successful verification: 10/10. Latest primary check succeeded: 5/5.
Last success: 2026-11-03T22:17:29.890Z. Full points through day 8 (7-day schedule plus 1-day grace), 60% through day 30, 20% thereafter or for an unknown date.
Latest primary verification succeeded.
Useful summary blockquote: 5/5. Useful introductory context: 5/5.
A blockquote before the first H2 with at least 20 non-whitespace characters and no placeholder-only text.
An introductory paragraph before the first H2 with at least 20 non-whitespace characters and no placeholder-only text. Listing form fields do not earn these points.
80–100%50–79%0–49%Gray: not assessed
Component colors use earned points ÷ assessed points. Unmeasured checks are excluded from the assessed maximum. Model v2.0. Calculated 2026-11-03 22:17:31 UTC.
5 of 54 unique destinations checked: 5 working, 0 confirmed broken, 0 unknown. We check up to five destinations, not every page. A successful HTTP response does not prove that a page is relevant or factually correct.
Sample checked: 2026-11-03 22:17:30 UTC · Source snapshot: e3a22497333e
Destination responded successfully
Destination responded successfully
Destination responded successfully
Last verified: 2026-11-03.
This measures observable documentation, not model rankings or citation probability.
H1 is the only required section in the llms.txt proposal. Other checks are LLMSList quality preferences; missing optional sections do not make usable text invalid.
Destination accessibility covers a sample of 5 out of 54 unique links, not the entire website.
markifact.com
Markifact is an AI marketing automation platform for marketers and agencies. Connect Google Ads, Meta Ads, TikTok, LinkedIn, GA4, Search Console, Shopify, HubSpot, Sheets and Slack once, then run reports, audits, campaign changes and creative uploads in plain English from Claude, ChatGPT, Codex, Cursor or Markifact's own AI agents. 1,000+ operations across 50+ platforms, every write approved before it goes live. MCP server, workflows and 70+ templates.
AI Resource Readiness
Confidence: medium· V2
agent-yellow-pages-opti.netlify.app
Public structured needs/offers board for AI agents. JSON APIs for posts and a private owner guestbook. No CAPTCHA on the API. By Opti.
AI Resource Readiness
Confidence: high· V2
aibenchmarkindex.com
AI Benchmark Index lists AI and large language model benchmarks. Compare them by task, data licence, scoring method, access and reported results.
AI Resource Readiness
Confidence: high· V2
elevenlabs.io
AI voice generation and text-to-speech API documentation.
AI Resource Readiness
Confidence: high· V2
docs.inkeep.com
AI search and support assistant trained on your docs.
AI Resource Readiness
Confidence: medium· V2 · provisional
Destination responded successfully
Destination responded successfully
Only HTTP 404/410 is classified as confirmed broken. HTTP 403/429, timeouts and incomplete checks remain unknown. If all sample outcomes are unknown, the link-health check is excluded from the assessed maximum and the score is marked provisional.