Hacker News: blackmanta

New comment by blackmanta in "I ran Gemma 4 as a local model in Codex CLI"

blackmanta — Mon, 13 Apr 2026 01:27:33 +0000

With a nvidia spark or 128gb+ memory machine, you can get a good speed up on the 31B model if you use the 26B MoE as a draft model. It uses more memory but I’ve seen acceptance rate at around 70%+ using Q8 on both models

New comment by blackmanta in "How the U.S. National Science Foundation enabled Software-Defined Networking"

blackmanta — Wed, 29 Oct 2025 23:29:55 +0000

Unless the goal of the backdoor is to redirect traffic flows through packet inspection devices that the attacker also controls, the decoupling of the control and data plane in SDN deployments requires a more creative, intricate solution to allow for wiretapping compared to traditional routers.

New comment by blackmanta in "Context engineering is sleeping on the humble hyperlink"

blackmanta — Sat, 25 Oct 2025 13:34:06 +0000

You can actually enable search in codex but adding the following to the config file

[tools] web_search = true

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Tue, 19 Aug 2025 00:07:20 +0000

I will look into adding this in a future update.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Tue, 19 Aug 2025 00:07:05 +0000

The graph functionality is stubbed but I will expose it in a future update. You can also use metadata and tags for similar things.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 19:22:34 +0000

I may explore this when I implement the vectordb implementation I started.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 18:09:44 +0000

While the examples and provided prompt lean toward code (since that's my personal use case), YAMS is fundamentally a generic content-addressed storage system.

I will attempt to run some small agents with custom prompts and report back.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 16:50:27 +0000

I have been using it for task tracking, research, and code search. When using CLI tools, I found that the LLM's were able to find code in less tool calls when I stored my codebase in the tool. I had to wrangle the LLMs to use the tool verse native rgrep or find.

I am also trying to stabilize PDF text extraction to improve knowledge retrieval when I want to revisit a paper I read but cannot remember which one it was. Most of these use cases come from my personal use and updates to the tool but I am trying to make it as general as possible.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 15:10:26 +0000

I stored the codebase for yams in the tool. The "blocks" are content-defined blocks/chunks, not filesystem blocks. They're variable-size chunks (typically 4-64KB) created using Rabin fingerprinting to find natural content boundaries. This enables deduplication across files that share similar content.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 14:48:28 +0000

I have not, but that is something I plan to do when I have time.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 14:44:04 +0000

I am working to improve the CLI tools to make getting this information easier but I have stored the yam repo in yams with multiple snapshots and metadata tags and I am seeing about 32% storage savings.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 14:24:44 +0000

The graph functionality is exposed through the retrieval functionality. I may improve this later but the idea was to maximize getting the best results when looking for stored data.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 14:21:34 +0000

The system is honestly tuned for storage efficiency not speed but these configurations are tunable and you can use the benchmarks as a reference for tuning. https://github.com/trvon/yams/blob/main/docs/benchmarks/perf...

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 05:25:56 +0000

The tool has built-in versioning. Each file gets a unique SHA-256 hash on storage (automatic versioning), you can update metadata to track version info, and use collections/snapshots to group versions together. I have been using the metadata to track progress and link code snippets.

New comment by blackmanta in "Show HN: Yet another memory system for LLMs"

blackmanta — Thu, 14 Aug 2025 04:43:23 +0000

That should be fixed now. It was a misconfiguration of CloudFlare SSL with GitHub Pages.

Show HN: Yet another memory system for LLMs

blackmanta — Thu, 14 Aug 2025 03:34:11 +0000

Built this for my LLM workflows - needed searchable, persistent memory that wouldn't blow up storage costs. I also wanted to use it locally for my research. It's a content-addressed storage system with block-level deduplication (saves 30-40% on typical codebases). I have integrated the CLI tool into most of my workflows in Zed, Claude Code, and Cursor, and I provide the prompt I'm currently using in the repo.

The project is in C++ and the build system is rough around the edges but is tested on macOS and Ubuntu 24.04.

Comments URL: https://news.ycombinator.com/item?id=44896489

Points: 165

# Comments: 46

Show HN: Leveraging Google ADK for Cyber Threat Intelligence

blackmanta — Sat, 28 Jun 2025 20:19:50 +0000

The project is in an early state. I just had to recently reset the graph data store, but I figured now is a good time to share my post and project.

The link is to my blog post, the tool is at https://umbrix.dev

Comments URL: https://news.ycombinator.com/item?id=44407845

Points: 4

# Comments: 1

Ex-OpenAI board member reveals what led to Sam Altman's brief ousting

blackmanta — Tue, 28 May 2024 23:13:33 +0000

Article URL: https://www.businessinsider.com/openai-board-member-details-sam-altman-lied-allegation-ousted-2024-5

Comments URL: https://news.ycombinator.com/item?id=40506582

Points: 785

# Comments: 805

New comment by blackmanta in "Google Messages RCS is blocked on rooted, custom ROM Android phones"

blackmanta — Sat, 02 Mar 2024 04:06:21 +0000

I have used Google messages with RCS on grapheneos.

New comment by blackmanta in "Microsoft Security-101: Open-Source curriculum"

blackmanta — Sat, 02 Mar 2024 04:02:56 +0000

Microsoft does take security pretty seriously with windows. There have been times they make mistakes, but historically, they are doing better now than they were in the early 2000s before Bill Gates sent a memo noting that they needed to take security seriously. You could argue they take security as serious as any major play like apple and google, seeing they touch a lot of private user data.

[1] https://news.microsoft.com/2000/12/07/gates-offers-new-techn...