惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Google DeepMind News
Google DeepMind News
Exploit-DB.com RSS Feed
Exploit-DB.com RSS Feed
Security Latest
Security Latest
P
Palo Alto Networks Blog
AWS News Blog
AWS News Blog
NISL@THU
NISL@THU
T
Threatpost
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
Latest news
Latest news
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
WordPress大学
WordPress大学
J
Java Code Geeks
P
Privacy International News Feed
阮一峰的网络日志
阮一峰的网络日志
S
Schneier on Security
博客园 - 聂微东
Project Zero
Project Zero
美团技术团队
Recent Commits to openclaw:main
Recent Commits to openclaw:main
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
Scott Helme
Scott Helme
I
Intezer
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
H
Hacker News: Front Page
S
Security @ Cisco Blogs
博客园 - 司徒正美
O
OpenAI News
Last Week in AI
Last Week in AI
L
LINUX DO - 热门话题
酷 壳 – CoolShell
酷 壳 – CoolShell
SecWiki News
SecWiki News
月光博客
月光博客
S
Security Affairs
The GitHub Blog
The GitHub Blog
P
Privacy & Cybersecurity Law Blog
S
Secure Thoughts
V
V2EX
S
Securelist
F
Fortinet All Blogs
W
WeLiveSecurity
D
Docker
博客园 - 三生石上(FineUI控件)
Simon Willison's Weblog
Simon Willison's Weblog
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
C
Cyber Attacks, Cyber Crime and Cyber Security
V
Visual Studio Blog
www.infosecurity-magazine.com
www.infosecurity-magazine.com
Webroot Blog
Webroot Blog
Engineering at Meta
Engineering at Meta

Hacker News - Newest: "AI"

AI can't read an investor deck AI as an attorney? Student uses ChatGPT, Gemini to sue UW over alleged racial discrimination Hacking MCP Servers in AI Systems – The Rug Pull: Tool Changes After Approval GitHub - MeepCastana/KubeezCut: Free Web based video editor GitHub - GenAI-Gurus/awesome-eu-ai-act: Curated tools, official sources, OSS, templates, and guides for EU AI Act compliance. Can AI judge journalism? A Thiel-backed startup says yes, even if it risks chilling whistleblowers Coming soon: 10 Things That Matter in AI Right Now DARPA built an AI to fact-check enemy weapons claims What explains heterogeneity in AI adoption? When AI Meets Muscle: Context-Aware Electrical Stimulation Promises a New Way to Guide Human Movements - Department of Computer Science AI Changed How We Build. It Did Not Change What Matters. Linux rules on using AI-generated code - Copilot is OK, but humans must take 'full responsibility for the… Meta spins up AI version of Mark Zuckerberg to engage with employees Code Mode: Let Your AI Write Programs, Not Just Call Tools | TanStack Blog GitHub - Delavalom/graft: Go framework for building AI agents. Type-safe tools, multi-provider (OpenAI, Anthropic, Gemini, Bedrock), zero vendor SDKs. India's TCS tops estimates, says new AI models did not dent services demand Gen Z's fading AI hype Strong feeling: we are in a folded AI reality GitHub - machinarii/total-recall-catalog: A reference catalog of latest knowledge retrieval, memory & RAG systems GitHub - mensfeld/code-on-incus: Give each AI agent its own isolated machine with root, Docker, and systemd. Active defense detects and stops threats automatically.. Quantization, LoRA, and the 8% Problem: Benchmarking Local LLMs for Production AI Iran war: We spoke to the man making Lego-style AI videos that experts say are powerful propaganda Powell, Bessent discussed Anthropic's Mythos AI cyber threat with major U.S. banks GitHub - immartian/bellamem: Persistent belief-graph memory for AI agents. Retrieves decisive context by importance — not recency, not RAG, not /compact. recursive-mode: The Repo-Native Operating System for AI Engineering After the attack on Sam Altman's home, will AI CEO's go on the offensive? The biggest advance in AI since the LLM Opus 4.6 vs GPT 5.4 One Prompt Unity World Generation Test “AI polls” are fake polls Client Challenge Can AI be a 'child of God'? Inside Anthropic's meeting with Christian leaders How to Switch AI Chatbots and Why You Might Want To GitHub - MattMessinger1/agentic_refund_guardrail: Safe refund policy layer for AI agents — Python + TypeScript. Same behavior, shared tests. Adam/papers/emergent_values_whitepaper.md at master · strangeadvancedmarketing/Adam Ask HN: How do you stop playing 20 questions with your AI coding tools How far can automation and AI support psychotherapy? - @theU GitHub - stagas/rtdiff: realtime git diff gui and AI-assisted commits A Mac Studio for Local AI — 6 Months Later A History of the Early Years of AI at the University of Edinburgh Why AI Coding Tools Still Feel Stuck on Localhost MSN AI Datacenters Are Becoming Strategic Targets twitter.com Penn Researchers Use AI to Surface Unreported GLP-1 Side Effects in Reddit Posts Show HN: MoodSense AI (ML and FastAPI and Gradio, Deployed on Hugging Face) Moodsense Ai - a Hugging Face Space by aman179102 AI models are terrible at betting on soccer—especially xAI Grok GitHub - xialeistudio/echoic GitHub - HimashaHerath/github-dev-wrapped: AI-powered weekly GitHub activity reports deployed to GitHub Pages GitHub - alejandrobalderas/claude-code-from-source: Architecture, patterns & internals of Anthropic's AI coding agent — reverse-engineered from source maps AI and Tech brief: Ireland ascendant GitHub - Titovilal/context0: Context0 - Never Surrender Training for a Marathon with an AI Coach: What Worked and What Didn't Cyber Pulse: Agentic Intel - Apps on Google Play I Built an AI PR Reviewer That Catches Bugs by Not Looking for Bugs Gen Z workers are so fearful AI will take their job they’re intentionally sabotaging their company’s AI rollout | Fortune How AI Is Reimagining the Game of Golf–For Both Players and Courses GitHub - nattergabriel/reseed: A CLI tool for managing and distributing agent skills across projects Is SVG the final frontier? My AI workflow evolved from prompts to a near-autonomous workflow MLSharp Help - 3DGS Viewer & Generator I put my cognitive field based AI's runtime on GitHub Is Numble the first AI-proof game? A3: Kubernetes for autonomous AI agent fleets | Emergent Principles Deepali Vyas ("The Elite Recruiter") GitHub - msmarkgu/RelayFreeLLM: A restful API designed to route user prompts to various AI model providers. Unionized ProPublica staff are on strike over AI, layoffs, and wages Unleashing the Advantage of Quantum AI We're heading for an AI-fueled 'dementia crisis,' brain scientist warns The AI-Assisted Breach of Mexico's Government Infrastructure [pdf] GitHub - stef41/lmscan: 🔍 Detect AI-generated text and fingerprint which LLM wrote it. Open-source GPTZero alternative. Zero dependencies, works offline. MSN GitHub - visionscaper/collabmem: Enabling long-term collaboration with Agentic AI - building up episodic and world model memory over time with in-context awareness We gave an AI a 3 year retail lease in SF and asked it to make a profit | Andon Labs AI Code is Hollowing Out Open Source, and Maintainers are Looking the Other Way What leaked "SteamGPT" files could mean for the PC gaming platform's use of AI AI is the boss at this retail store. What could go wrong? GitHub - Wuzu11517/agentic-proxy: Local proxy meant to help reduce With Drones, Geophysics and ArtificiaI Intelligence, Researchers Prepare to Do Battle Against Land Mines A Single Operator, Two AI Platforms, Nine Government Agencies: The Full Technical Report 在 Steam 上购买 FriedrichAI: Offline AI 立省 10% GitHub - inevolin/resume-cli: Hit Claude usage limits? Resume any AI coding session elsewhere. Switch tools at zero friction. GitHub - atripati/ark: AI Runtime Kernel — a context operating system for AI agents. Eliminates tool bloat, loads only what’s needed, and gives LLMs their reasoning space back. How to Build a Secure AI PR Reviewer with Claude, GitHub Actions, and JavaScript This Startup Wants You to Pay Up to Talk With AI Versions of Human Experts Intel Arc Pro B70 Brings 32GB VRAM to Local AI for $949 WordPress 7.0: The Good, the AI, and the Still Missing AI on the couch: Anthropic gives Claude 20 hours of psychiatry IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures AI Agents Know About Supabase. They Don't Always Use It Right. The history and future of AI at Google, with Sundar Pichai Inside an AI‑enabled device code phishing campaign How Meta Used AI to Map Tribal Knowledge in Large-Scale Data Pipelines AI for Systems: Using LLMs to Optimize Database Query Execution Forecasting the Economic Effects of AI Introducing Tinker: Play with AI, bring your ideas to life AI sheds light on an ancient gaming mystery People really hate AI but not as much as Iran—or Democrats | Fortune What is an AI Product Engineer? Phoebe Gates wants her $185 million AI startup to succeed with 'no ties to my privilege or my last name': 'I have a chip on my shoulder' | Fortune
GitHub - vassiliylakhonin/agenda-intelligence-md: A markdown protocol for AI agents that analyze public agenda instead of summarizing it badly.
vassilbek · 2026-04-30 · via Hacker News - Newest: "AI"

Stars Forks Issues Last Commit Release v0.4.0 License

A drop-in markdown cognition layer for AI agents that need to analyze public agenda instead of summarizing it badly.

Most agent-written news analysis has the same problem: it tells you what happened, adds polished implications, and stops before the output changes any decision.

Agenda-Intelligence.md gives agents a stricter workflow and an optional reasoning-memory layer called AnalysisBank:

Fact → Assessment → Assumption → Unknown → Scenario → Indicator to watch

Use it when an agent needs to reason about policy, geopolitics, regulation, sanctions, trade, energy, elections, conflicts, or market-moving public agenda.


What's new in v0.4.0

v0.4.0 adds a general Source Acquisition Layer.

Reasoning is not enough. Agents need to know which source types are required before making claims about sanctions, regulation, elections, conflict, energy, trade, financial markets, AI/technology, or regional risk.

Added:

  • SOURCE_POLICY.md — source discipline rules;
  • source-taxonomy.json — machine-readable source types;
  • source-requirements/*.json — source plans by agenda category;
  • schemas/evidence-pack.schema.json — evidence pack contract;
  • examples/source/evidence-pack.json — valid example;
  • CLI commands: source-types, list-source-packs, source-plan, validate-evidence.

What's new in v0.3.0

v0.3.0 turns Agenda-Intelligence.md into a more agent-first package.

Added:

  • agent-manifest.json for machine-readable discovery;
  • JSON schemas for agenda briefs, memory cards, lens manifests, and signal classifications;
  • scripts/agenda_intelligence.py CLI for agents and humans;
  • MCP.md sketch for future MCP tools;
  • examples/agenda-brief.json for schema validation.

What's new in v0.2.0

v0.2.0 adds AnalysisBank, a ReasoningBank-inspired memory layer for agenda-analysis agents.

Instead of only giving an agent a protocol, AnalysisBank lets the project store short reusable reasoning memories from successful and failed outputs:

weak output → identify failure pattern → write memory card → retrieve next time
strong output → extract reusable reasoning pattern → write memory card → retrieve next time

It also adds a lightweight eval harness:

python3 scripts/eval_before_after.py

Current test results:

eu-ai-act.md:          before 3/16  → after 14/16
red-sea-shipping.md:  before 1/16  → after 13/16
sanctions-routing.md: before 2/16  → after 14/16

What this does

It pushes an agent to answer better questions:

  • Is this noise, weak signal, signal, structural shift, or trigger event?
  • What actually changed?
  • Who gains or loses leverage?
  • Which incentives shifted?
  • What is still unknown?
  • What would confirm or falsify this view?
  • What should be watched next?

The goal is not longer analysis. The goal is less filler and more decision value.


How to use it

Copy or reference the markdown files in your agent setup:

skills/agenda-intelligence/references/analysis-protocol.md
skills/agenda-intelligence/references/agenda-triage.md
skills/agenda-intelligence/references/evidence-discipline.md
skills/agenda-intelligence/references/output-patterns.md
analysis-bank/README.md

For repo-level agent instructions, link the base protocol from your AGENTS.md, system prompt, retrieval layer, or tool-specific skill wrapper.

Example instruction:

Before analyzing public agenda, use Agenda-Intelligence.md.
Do not summarize by default. Classify the signal, identify what changed, separate fact from assessment, name the main uncertainty, and end with watch-next indicators.

The repository also includes an OpenClaw-compatible skill wrapper, but the useful part is plain markdown and portable.


Built for agents

Agenda-Intelligence.md is designed to be consumed by agents, not just read by humans.

Agents can:

  • discover the package through agent-manifest.json;
  • load the entrypoint file with a stable path;
  • select regional and sector lenses programmatically;
  • validate structured outputs against JSON schemas;
  • run a lightweight CLI;
  • score before/after examples with the eval harness;
  • store reusable reasoning memories in AnalysisBank.

Agent-first files:

agent-manifest.json
schemas/agenda-brief.schema.json
schemas/memory-card.schema.json
schemas/lens-manifest.schema.json
schemas/signal-classification.schema.json
scripts/agenda_intelligence.py
SOURCE_POLICY.md
source-taxonomy.json
source-requirements/*.json
MCP.md

CLI examples:

python3 scripts/agenda_intelligence.py manifest
python3 scripts/agenda_intelligence.py list-lenses
python3 scripts/agenda_intelligence.py get-lens regional eu
python3 scripts/agenda_intelligence.py get-protocol entrypoint
python3 scripts/agenda_intelligence.py validate-brief examples/agenda-brief.json
python3 scripts/agenda_intelligence.py source-plan technology-ai
python3 scripts/agenda_intelligence.py validate-evidence examples/source/evidence-pack.json
python3 scripts/agenda_intelligence.py score

10-second demo

Without Agenda-Intelligence.md:

Companies should monitor developments and prepare for possible regulatory changes.

With Agenda-Intelligence.md:

Watch for regulator guidance, first enforcement action, compliance deadline, and company product redesigns. Treat this as a signal until those indicators appear.

The difference is not style. It is decision value.


Copy-paste setup

Fastest path: copy Agenda-Intelligence.md into your repo next to AGENTS.md.

Then add this to your agent instructions:

## Agenda analysis

When analyzing public agenda, news, policy, regulation, sanctions, geopolitics, trade, elections, conflicts, markets, or strategic risk, follow `Agenda-Intelligence.md`.

Do not summarize by default. Classify the signal, identify what changed, separate fact from assessment, name uncertainty, and end with watch-next indicators.

Load it conditionally. Do not add it to every task.

For deeper setups, also copy the relevant reference files from skills/agenda-intelligence/references/.


Default output shape

**Bottom line:** ...
**Signal classification:** noise / weak signal / signal / structural shift / trigger event
**What changed:** ...
**Why it matters:** ...
**Who is affected:** ...
**Main uncertainty:** ...
**Scenarios:** ...
**Watch next:** ...

Before / after examples

The repo includes concrete examples showing the failure mode this file is meant to fix.

Without the protocol:

recap → generic implications → “monitor developments”

With Agenda-Intelligence.md:

signal classification → what changed → affected actors → uncertainty → scenarios → watch-next indicators

Examples:


What it is good for

  • research and news agents;
  • policy and geopolitical agenda tracking;
  • sanctions and compliance monitoring;
  • trade and regulatory risk briefs;
  • founder/investor operating-context notes;
  • NGO/donor context monitoring;
  • election and diplomatic signal analysis;
  • red-team checks on confident narratives.

What it is not

  • not legal advice;
  • not investment advice;
  • not an intelligence-certainty machine;
  • not a news summarizer;
  • not a replacement for source verification.

If live verification was not performed, the agent should say so.


How it relates to AGENTS.md

AGENTS.md tells an agent how to operate. Agenda-Intelligence.md tells an agent how to reason about public agenda.

Use AGENTS.md globally. Use Agenda-Intelligence.md conditionally when the task involves news, policy, regulation, sanctions, geopolitics, trade, elections, conflicts, markets, or strategic risk.

In practice, it can sit next to the usual agent files:

AGENTS.md      = operating rules
SOUL.md        = voice and stance
TOOLS.md       = tool discipline
IDENTITY.md    = agent identity
USER.md        = user preferences
HEARTBEAT.md   = proactive behavior
MEMORY.md      = durable context
Agenda-Intelligence.md = public-agenda reasoning protocol

Minimal AGENTS.md hook:

## Agenda analysis

When analyzing public agenda, policy, regulation, sanctions, geopolitics, trade, elections, conflicts, markets, or strategic risk, follow `Agenda-Intelligence.md`.

Do not summarize by default. Classify the signal, identify what changed, separate fact from assessment, name uncertainty, and end with watch-next indicators.

The important part is conditional loading. Do not spend context on agenda analysis rules when the task is ordinary coding, writing, or personal assistance.


Source Acquisition Layer

Agenda-Intelligence.md separates reasoning from evidence.

Before writing a high-stakes or current brief, an agent should generate a source plan:

Task → source requirement category → required evidence → unsupported claims → brief

Available source requirement categories:

  • sanctions
  • regulation
  • elections
  • conflict-security
  • energy
  • trade
  • financial-market
  • technology-ai
  • regional-risk

CLI examples:

python3 scripts/agenda_intelligence.py source-types
python3 scripts/agenda_intelligence.py list-source-packs
python3 scripts/agenda_intelligence.py source-plan technology-ai
python3 scripts/agenda_intelligence.py validate-evidence examples/source/evidence-pack.json

If live retrieval fails, the agent should say so and downgrade evidence mode instead of pretending the brief is source-backed.


AnalysisBank

AnalysisBank is the ReasoningBank-inspired layer for Agenda-Intelligence.md.

The base protocol tells an agent how to analyze public agenda. AnalysisBank helps it improve across tasks by storing compact reasoning memories from both good and bad outputs.

Current memory cards include:

  • vague monitoring → concrete indicators;
  • overconfident sanctions upgrades → evidence thresholds;
  • EU rhetoric treated as law → institutional-path check;
  • sanctions routing → mechanism-first signal classification.

Memory format:

Trigger → Pattern → Better reasoning → Apply when → Do not apply when → Watch indicators → Example rewrite

Eval harness:

python3 scripts/eval_before_after.py

The eval checks that after examples score higher than generic before examples on signal classification, actor specificity, uncertainty, falsifiability, watch-next indicators, and decision value.


Regional lens packs

Agenda-Intelligence.md can be extended with lightweight regional thinking layers. These are not full specialist skills; they are portable checklists that help any agent reason better about a specific region.

Available lens packs:

  • Central Asia + Caspian — sanctions routing, corridor politics, Caspian chokepoints, banking/payment exposure, state leverage, energy, minerals, and regional political economy.
  • Middle East — escalation risk, energy flows, maritime chokepoints, sovereign capital, sanctions exposure, normalization, and regional power competition.
  • European Union — regulation, sanctions, trade defense, digital rules, climate policy, market access, coalition politics, and enforcement risk.

Use the base protocol first, then add the regional lens when the agenda item has a clear regional connection.


Sector lens packs

Sector lenses add domain-specific checks for high-risk agenda areas. They are not legal, financial, or technical advice; they are reasoning checklists for agents.

Available sector packs:

  • Sanctions — designations, enforcement, export controls, routing, ownership/control, financial channels, licenses, and compliance exposure.

Use the base protocol first, then add the sector lens when the agenda item has a clear domain connection.


Relationship to global-think-tank-analyst

Agenda-Intelligence.md is the lightweight, portable agenda-analysis protocol. It is for any AI agent that needs to stop summarizing news and start identifying signal, uncertainty, scenarios, and watch-next indicators.

For full policy-risk memos, use global-think-tank-analyst. That repository is the deeper OpenClaw/Codex analyst for geopolitical, sanctions, trade, regulatory, and strategic-risk memos.

Use them together like this:

Agenda-Intelligence.md = small universal protocol for agenda triage
global-think-tank-analyst = full memo skill for decision-ready policy risk analysis

Repository structure

Agenda-Intelligence.md
ADOPTION.md
agent-manifest.json
MCP.md
SOURCE_POLICY.md
source-taxonomy.json
source-requirements/
  sanctions.json
  regulation.json
  elections.json
  conflict-security.json
  energy.json
  trade.json
  financial-market.json
  technology-ai.json
  regional-risk.json
analysis-bank/
  README.md
  MEMORY_FORMAT.md
  failures/
  successes/
  prompts/
skills/agenda-intelligence/
  SKILL.md
  references/
    analysis-protocol.md
    agenda-triage.md
    evidence-discipline.md
    output-patterns.md
    regional/
      central-asia-caspian.md
      middle-east.md
      eu.md
    sector/
      sanctions.md
examples/
  compact-brief.md
  red-team-brief.md
  central-asia-caspian-brief.md
  middle-east-brief.md
  eu-brief.md
  sector/
    sanctions-brief.md
  before-after/
    eu-ai-act.md
    red-sea-shipping.md
    sanctions-routing.md
schemas/
  agenda-brief.schema.json
  memory-card.schema.json
  lens-manifest.schema.json
  signal-classification.schema.json
  evidence-pack.schema.json
scripts/
  agenda_intelligence.py
  validate.py
  eval_before_after.py
llms.txt

Design principle

Keep the loaded context small.

SKILL.md is only a wrapper. The deeper markdown files are pulled or copied only when needed:

  • analysis-protocol.md — how the agent should think;
  • agenda-triage.md — how to classify developments;
  • evidence-discipline.md — how to handle uncertainty and sources;
  • output-patterns.md — ready-to-use brief formats;
  • regional/central-asia-caspian.md — regional lens for Central Asia + Caspian agenda analysis;
  • regional/middle-east.md — regional lens for Middle East agenda analysis;
  • regional/eu.md — regional lens for European Union agenda analysis;
  • sector/sanctions.md — sector lens for sanctions and export-control agenda analysis.

License

MIT