frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Tako AI – Agent for Okta With Natural language (zero hallucination)

https://github.com/fctr-id/okta-ai-agent
1•danFctr•1h ago
Hi HN,

Every week I watched Okta admins burn hours answering ad-hoc questions from security teams: "Who has access to Salesforce?", "Find all contractors with GitHub access who haven't used MFA in 30 days." The answers always involved the same painful loop: dig through a slow web console, chain API calls, correlate CSVs, write throwaway Python scripts. Repeat next week.

I spent 12 months building Tako AI to fix this. You ask a question in plain English, it returns verified data.

GitHub: https://github.com/fctr-id/okta-ai-agent

THE ONE RULE: Zero hallucinations.

In identity and access management, a wrong answer is worse than no answer. If an AI tells your CISO a contractor doesn't have admin access when they actually do, that's a security incident. Tako never "predicts" an answer. It writes the code to find the answer, executes it, and returns the raw result. Ask the same question twice, you get the same data.

THE HARDEST PROBLEM: Scaling to 107+ API endpoints

Most AI agents break down past 10-20 tools. They hallucinate parameters, call wrong endpoints, invent fields that don't exist. We went through five architecture rewrites over 12 months.

Each iteration: new LLM drops (GPT-4, Claude 3.5), we rebuild the agent, hit context limits, watch it snowball into gibberish. The breakthrough wasn't bigger context windows — it was precise context engineering. Instead of cramming 107 endpoint definitions into a prompt, the agent dynamically discovers the right spec for the task at hand. It reads a custom JSON API documentation file for the specific endpoint it needs, constructs validated requests, executes them. No hardcoded tools per endpoint. We're adding full CRUD operations next.

HOW IT WORKS:

Multi-agent architecture based on ReAct (Reasoning + Acting). Each agent has a narrow job:

• Router: analyzes your question, decides local cache vs live API • SQL Agent: queries local SQLite cache for bulk data (10k users in milliseconds vs minutes via API) • API Agent: handles live Okta calls • Synthesis Agent: merges everything into final verified report

The API Agent has a self-healing loop that surprised us. When generated code fails — wrong parameter name, rate limit hit, API schema changed — it traps the stack trace, feeds the error back to the LLM with context, and rewrites the code. We've seen it recover from Okta API changes we didn't even know happened yet.

PRIVACY & SECURITY:

Runs 100% locally in Docker. You bring your own LLM keys (OpenAI, Anthropic, Gemini, or Ollama for fully offline). Your employee PII never leaves your machine.

READ-ONLY by design. All generated Python and API code runs in a sandboxed environment. Every execution is automatically verified against security patterns before running — code is logged and available for audit, but you don't manually approve each query.

WHAT'S NEXT:

We see this as a platform, not just an Okta tool. The pattern (local cache + live ReAct agent + self-healing code execution) generalizes to any SaaS API. Google Workspace, Slack, Workday — same architecture, different spec files. Working on write operations with human-in-the-loop approval next.

What would you want AI agents to actually do for you in 2026? Where do you see this tech going beyond chatbots?

—Dan

Launch HN: Omnara (YC S25) – Run Claude Code and Codex from Anywhere

3•kmansm27•1m ago•0 comments

YouTube Launches on Apple Vision Pro

https://www.macrumors.com/2026/02/12/youtube-app-apple-vision-pro/
1•tosh•1m ago•0 comments

Browser-based tool for generating songs from text

https://texttosong.ai/
1•pekable•3m ago•1 comments

AI Coding Agents Will 10x Your Datadog Bill

https://oneuptime.com/blog/post/2026-02-12-ai-coding-agents-will-10x-your-datadog-bill/view
1•ndhandala•3m ago•0 comments

AI Conversations Aren't Privileged

https://twitter.com/mpeltz/status/2021778562328482231
1•delichon•4m ago•0 comments

Show HN: Rebuilding My First Startup with Claude Agent SDK

https://laminar.sh/blog/2026-02-10-rebuilding-my-first-startup-as-an-ai-agent
1•samkom•4m ago•0 comments

Telegram CEO condemns new restrictions in Russia as citizens turn to VPNs

https://www.techradar.com/vpn/vpn-privacy-security/telegram-ceo-condemns-new-restrictions-in-russ...
1•maxloh•4m ago•0 comments

Study: Tracking devices on loose clothing provide more accurate movement data

https://www.kcl.ac.uk/news/beyond-the-fitbit-why-your-next-health-tracker-might-be-a-button-on-yo...
1•giuliomagnifico•4m ago•0 comments

Show HN: TinyFish Web Agent (82% on hard tasks vs. Operator's 43%)

https://www.tinyfish.ai/blog/mind2web
5•gargi_tinyfish•5m ago•2 comments

Agents and Identity – Navigating What We Can't Predict [audio]

https://packetpushers.net/podcasts/the-cloud-gambit/tcg068-agents-and-identity-navigating-what-we...
1•mooreds•5m ago•0 comments

Announcing TypeScript 6.0 Beta

https://devblogs.microsoft.com/typescript/announcing-typescript-6-0-beta/
1•maxloh•6m ago•0 comments

Beyond SAST: Using Gemini to Orchestrate Semantic Source Reviews

https://ciex-software.com/llm-appsec.html
1•wglb•6m ago•0 comments

Most-Viewed People on Wikipedia in 2025 (Catalyst Events and Social Memory)

https://blog.wolfram.com/2026/02/12/most-viewed-people-on-wikipedia-in-2025-how-catalyst-events-i...
2•soofy•6m ago•0 comments

Coding Agents Meet Distributed Reality

https://jhellerstein.github.io/blog/codegen-reality/
1•shadaj•7m ago•0 comments

Shut Up: Comment Blocker

https://rickyromero.com/shutup/
2•mefengl•10m ago•0 comments

Why Germany is racing to rebuild its army

https://www.theguardian.com/news/audio/2026/jan/26/why-germany-is-racing-to-rebuild-its-army
3•PaulHoule•11m ago•0 comments

GitHub Feb 9th outage: Incident Report

https://www.githubstatus.com/incidents/smf24rvl67v9
4•whyleyc•12m ago•1 comments

The Redundancy Paradox

https://www.mihirdeshpande.com/posts/redundancy_paradox
2•mihirrd•12m ago•0 comments

ai;dr

https://www.0xsid.com/blog/aidr
2•ssiddharth•12m ago•0 comments

Norway's former PM charged with gross corruption over Epstein links

https://www.bbc.com/news/articles/c5yqr8eggvwo
2•belter•13m ago•1 comments

Discord says 'vast majority' of users won't see its new age verification setup

https://www.theverge.com/tech/876575/discord-age-verification-vast-majority-users-inference
2•Alupis•13m ago•1 comments

More lessons from 14 years at Google

https://addyo.substack.com/p/14-more-lessons-from-14-years-at
3•ingve•13m ago•0 comments

Something Bigger Is Being Ignored

https://medium.com/@shankhadey/something-bigger-is-being-ignored-a43c6ee55c55
2•shubhodey•15m ago•0 comments

Review Papers May Matter More Than Experiments

https://evanwarfel.substack.com/p/your-understanding-of-the-scientific
2•YossarianFrPrez•15m ago•1 comments

Extracting and Analyzing Apple sysdiagnose Logs (2025)

https://blog.elcomsoft.com/2025/06/extracting-and-analyzing-apple-unified-logs/
1•walterbell•15m ago•0 comments

Who Is Paying for the 2025 U.S. Tariffs?

https://libertystreeteconomics.newyorkfed.org/2026/02/who-is-paying-for-the-2025-u-s-tariffs/
2•lysace•15m ago•1 comments

AMA launches independent vaccine review after CDC criticism

https://medicalxpress.com/news/2026-02-ama-independent-vaccine-cdc-criticism.html
1•bikenaga•16m ago•0 comments

Analyzing Container Filesystem Isolation for Multi-Tenant Workloads

https://medium.com/@epappas/i-am-breaking-my-head-in-analyzing-container-filesystem-isolation-for...
1•hevalon•17m ago•0 comments

Pete Budwit

https://juanapril.substack.com/p/the-legal-doctrine-protecting-fascist
1•4slider•17m ago•0 comments

Substack CEO Chris Best: 'You can't define success as 'no one is ever mad at us'

https://speedrun.substack.com/p/substack-ceo-chris-best-you-cant
1•7777777phil•19m ago•0 comments