Feed aggregator

Article URL: https://app.quantra.io/

Comments URL: https://news.ycombinator.com/item?id=47301433

Points: 1

# Comments: 0

Article URL: https://worldconflictmonitor.com/

Comments URL: https://news.ycombinator.com/item?id=47301404

Points: 1

# Comments: 1

Article URL: https://essaysbyandy.substack.com/p/drink-the-radioactive-gatorade

Comments URL: https://news.ycombinator.com/item?id=47301399

Points: 1

# Comments: 0

With the recent incidents (DataTalks database wipe by Claude Code, Replit agent deleting data during code freeze), it's clear that running AI agents in production without observability is risky.

Common failure modes I've seen: no visibility into what the agent did step-by-step, surprise LLM bills from untracked token usage, risky outputs going undetected, and no audit trail for post-mortems.

I've been building AgentShield (https://useagentshield.com) — an observability SDK for AI agents. It does execution tracing, risk detection on outputs, cost tracking per agent/model, and human-in-the-loop approval for high-risk actions. Plugs into LangChain, CrewAI, and OpenAI Agents SDK with a 2-line integration.

Curious what others are using. Rolling your own monitoring? LangSmith? Langfuse? Or just hoping for the best?

Comments URL: https://news.ycombinator.com/item?id=47301395

Points: 1

# Comments: 0

The pipeline takes a trending topic and produces a fully rendered ~22 min comedy podcast episode with three AI characters (PRODUCER, CRITIC, DUMBASS). Everything is automated: premise ideation, research, outline generation, script writing with a writers' room (punch-up passes, verification gates), voice synthesis via ElevenLabs, music bed mixing, and publishing to Spotify.

Stack: Temporal for durable workflow orchestration, Gemini for script generation, gollem (https://github.com/fugue-labs/gollem) agents with structured outputs and automatic validation, Postgres + Apache AGE for graph queries, Qdrant for vector search, ElevenLabs for multi-voice dialogue synthesis.

Each episode goes through ~10 beats, each independently written and verified for grounding. There's a verifier gate that checks factual claims, forbidden phrases, and character voice consistency before anything gets rendered. The whole thing runs with a single slash command and takes about 2 hours end-to-end.

24 episodes in so far. This one's about a character who uses AI to write a book in a weekend and declares himself a literary genius.

Comments URL: https://news.ycombinator.com/item?id=47301386

Points: 1

# Comments: 0

Article URL: https://github.com/sylvinus/agent-vm

Comments URL: https://news.ycombinator.com/item?id=47301380

Points: 1

# Comments: 1

Article URL: https://ucode.mein.io/

Comments URL: https://news.ycombinator.com/item?id=47301348

Points: 1

# Comments: 0

Article URL: https://github.com/mnafees/openports

Comments URL: https://news.ycombinator.com/item?id=47301337

Points: 2

# Comments: 0

Hello HN :)

Promptinel is a security scanner for prompts. I think the world needs something like this, because prompts are basically executable artifacts and we need to treat them as such. I want to identify as many bad things as I can before a prompt gets anywhere near runtime. Especially when supplying a common prompt repository for friends or colleagues, or when downloading skills from the internet. You can read more about my motivation in the project's readme: https://github.com/CunningFatalist/promptinel?tab=readme-ov-...

Promptinel finds various attack patterns in prompts, for example:

- prompt override and role spoofing patterns

- download-and-execute chains

- template execution and network fetch behavior

- secret exfiltration intent

- invisible Unicode and obfuscation tricks

- local sensitive file references

I'm really looking forward to feedback from people, who have built linters or security scanners before. Or just feedback from Go people, really. I'm mainly a PHP and TypeScript dev and use Go in my free time, because I find it fun and love the philosophy behind it.

Comments URL: https://news.ycombinator.com/item?id=47301327

Points: 1

# Comments: 0

Article URL: https://pablooliva.de/the-closing-window/the-workplace-blind-spot/

Comments URL: https://news.ycombinator.com/item?id=47301299

Points: 1

# Comments: 0

I built Exolio AI to help teachers detect AI-generated student work. It uses a custom RoBERTa model trained on human and AI text. One free check with no account needed. Happy to answer any questions about how it works.

Comments URL: https://news.ycombinator.com/item?id=47301283

Points: 1

# Comments: 0

Article URL: https://www.madebymikal.com/interpreting-whiteout-files-in-docker-image-layers/

Comments URL: https://news.ycombinator.com/item?id=47301279

Points: 2

# Comments: 0

Article URL: https://www.shiptraffic.net/2016/04/Gibraltar-strait-ship-traffic.html

Comments URL: https://news.ycombinator.com/item?id=47301277

Points: 2

# Comments: 0

Article URL: https://www.vspvision.com/newsroom/eye-health-and-wellness/vision-care/new-research-from-vsp-vision-care-and-workplace-intelligence-finds-desk-workers-spend-nearly-100-hours-a-week-on-screens

Comments URL: https://news.ycombinator.com/item?id=47301275

Points: 12

# Comments: 6

Article URL: https://www.seattletimes.com/seattle-news/times-watchdog/seattle-womans-911-calls-reveal-gaps-in-ambulance-service/

Comments URL: https://news.ycombinator.com/item?id=47301264

Points: 10

# Comments: 1

Article URL: https://thenewstack.io/skills-vs-mcp-agent-architecture/

Comments URL: https://news.ycombinator.com/item?id=47301253

Points: 2

# Comments: 0

Article URL: https://basicappleguy.com/basicappleblog/lil-finder-guy

Comments URL: https://news.ycombinator.com/item?id=47301241

Points: 2

# Comments: 0

Article URL: https://github.com/Rabrg/artificial-life

Comments URL: https://news.ycombinator.com/item?id=47301233

Points: 3

# Comments: 0

Article URL: https://apps.apple.com/se/app/reimagen-ai-music-to-art/id6757844075

Comments URL: https://news.ycombinator.com/item?id=47301224

Points: 1

# Comments: 0