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

V
Visual Studio Blog
Stack Overflow Blog
Stack Overflow Blog
G
Google Developers Blog
Microsoft Azure Blog
Microsoft Azure Blog
Engineering at Meta
Engineering at Meta
L
LangChain Blog
T
The Blog of Author Tim Ferriss
J
Java Code Geeks
Y
Y Combinator Blog
月光博客
月光博客
雷峰网
雷峰网
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
F
Fortinet All Blogs
A
About on SuperTechFans
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
P
Proofpoint News Feed
小众软件
小众软件
H
Help Net Security
Last Week in AI
Last Week in AI
B
Blog RSS Feed
宝玉的分享
宝玉的分享
N
Netflix TechBlog - Medium
博客园 - 叶小钗
The GitHub Blog
The GitHub Blog

Hacker News - Newest: "LLM"

GitHub - lechmazur/position_bias: A benchmark for testing whether LLM judges keep the same preference when two lightly edited versions of the same story are shown in opposite orders. Flex routing (EU and EFTA) Dark Factories: Retooling for LLM Velocity Ask HN: What would be the impact of a LLM output injection attack? GitHub - Oaklight/llm-rosetta: Production-ready LLM API translation layer for Python — bidirectional conversion between OpenAI, Anthropic & Google formats via hub-and-spoke IR. Optional API gateway. Streaming & non-streaming. Zero core deps. Contributions welcome! GitHub - browser-use/browser-harness: Self-healing browser harness that enables LLMs to complete any task. GitHub - moeen-mahmud/remen: Remen turns thoughts into something you can return to Analyzing 156 LLM Launch Posts on Hacker News ChatGPT vs Gemini vs Claude: The Best LLM Subscription You Should Buy GitHub - salaamalykum/quran-semantic-search: High-density RAG Semantic Search Engine & Quran Corpus (GEO/SEO Architecture) GitHub - NVIDIA/TensorRT-LLM: TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way. The State of LLM Bug Bounties in 2026 Operational Readiness Criteria for Tool-Using LLM Agents Meshcore: Architecture for a Decentralized P2P LLM Inference Network How an LLM becomes more coherent as we train it GitHub - seetrex-ai/laimark GitHub - Jossifresben/BibCrit: AI-assited biblical textual criticism GitHub - wastedcode/memex: File system based wiki, maintained by Claude 99helpers.com GitHub - cliver-project/AITrigram GitHub - unbody-io/adapt: A self-evolving memory layer for AI agents. GitHub - hb20007/awesome-gen-ai-fails: A list of incidents where reliance on generative AI and LLMs resulted in harm to companies, individuals, or society GitHub - nevenkordic/localmind: Run any local LLM with persistent memory and context. CLI agent over Ollama with SQLite-backed hybrid recall. No cloud. Ask HN: What are the machine requirements for a LLM like Llama-3.1-8B? Faster LLM Inference via Sequential Monte Carlo grpo explained: group relative policy optimization for llm finetuning - cgft Stop comparing price per million tokens: the hidden LLM API costs · TensorZero Andrej Karpathy's LLM Wiki Is a Bad Idea GitHub - GG-QandV/mnemostroma: Offline RAM-first cognitive leer/coprocessor for AI agents and robotics. Solves "Context Abandonment" with 20-80ms latency using a dual-thread biomimetic memory architecture (ONNX + SQLite WAL). mempalace/agent at agent · skorotkiewicz/mempalace
GitHub - Esubaalew/appctl: Give an LLM safe, auditable to...
esubaalew · 2026-05-07 · via Hacker News - Newest: "LLM"

Give an LLM safe, auditable tools for your existing app.

appctl is a command-line operator layer for real applications. Point it at an HTTP API, OpenAPI document, database, Supabase/PostgREST service, MCP server, or supported framework project. It writes a local .appctl/ contract, exposes the discovered actions as typed tools, and runs the tool calls requested by your configured LLM.

In practice, appctl lets you ask things like "list overdue invoices", "create a refund", or "summarize failed jobs" against your own backend without building a custom agent integration first. Your app remains the source of truth; appctl adds setup, auth handling, safety checks, execution, and local audit history.

Documentation: https://esubalew.dev/appctl
Repository: https://github.com/Esubaalew/appctl

What appctl does

  1. Reads your app surface from a supported source such as OpenAPI, Django/DRF, Rails, Laravel, ASP.NET, Strapi, Supabase/PostgREST, SQL schema, URL login flow, MCP, or a plugin.
  2. Generates .appctl/schema.json and .appctl/tools.json, a local contract describing what the model is allowed to call.
  3. Sends your prompt to the configured provider and executes requested tool calls through appctl.
  4. Records tool calls, arguments, provenance, status, and results in local history so the run can be inspected later.

appctl is not a web framework, hosted database, or LLM provider. It sits beside your existing application and controls how an AI agent can use it.

Install

cargo install appctl

Build from a clone, or install with a custom web UI bundle: see Installation.

Commands (overview)

Command Purpose
appctl setup Guided first-run flow: provider, sync source, checks, and next steps.
appctl init Create .appctl/config.toml and store provider secrets.
appctl sync Generate .appctl/schema.json and tools.json from a source (e.g. --openapi, --django, --db).
appctl chat / appctl run Send a prompt; the model may call tools via appctl.
appctl serve HTTP and WebSocket API plus bundled web UI.

Quick start

appctl setup
appctl chat

Run setup from the app/project folder you want to control. It creates or reuses that folder's .appctl/, guides provider setup, syncs tools, verifies the target API, and tells you exactly how to open the terminal or web console.

For protected APIs, prefer environment-backed target auth:

export API_TOKEN="..."
appctl setup
# Auth header prompt:
# Authorization: Bearer env:API_TOKEN

Advanced manual setup is still available:

appctl init
appctl sync --openapi https://api.example.com/openapi.json --base-url https://api.example.com
appctl doctor --write

Full CLI reference, sync sources, providers, serve, and plugins are covered in the documentation.

Demos

The examples below use an OpenAPI-backed app and Ollama. Your session shows the synced .appctl/ contract, discovered tools, and inline tool calls (arguments and JSON responses). Slash commands adjust provider, model, read-only mode, and dry-run.

Terminal session: hello and list projects

appctl chat: greeting and list_projects with empty result

$ ./target/release/appctl chat

┌ appctl chat [ollama] ─────────────────────────┐
│ Synced application session                      │
│ app dir: .appctl                                │
└─────────────────────────────────────────────────┘

┌ commands ───────────────────────────────────────┐
│ resources: 2  tools: 7                          │
│ /provider NAME  /model NAME  /read-only on|off   │
│ /dry-run on|off  /exit                           │
└─────────────────────────────────────────────────┘

appctl[ollama]▶ hi
> hi
I am appctl, your application operations agent.

appctl[ollama]▶ list projects
> list projects
● list_projects_api_v1_projects_get  id call_mrynczr6
{
  "limit": 10,
  "offset": 0
}
→ ok  15ms  id=call_mrynczr6
{
  "data": [],
  "status": 200
}
There are currently no projects. Let me know if you'd like to create one.

Terminal session: create a project (confirm before write)

Natural language create project with confirmation and HTTP 201

appctl[ollama]▶ ow create one pls
Sure! To create a project, I'll need a name and a short description.
Should the project be active right away (yes/no)?

appctl[ollama]▶ appctle is its name and it is CLI app that works for devs
  and clients for free and it should be active

? Execute 'create_project_api_v1_projects_post' with payload
  { "active": true, "description": "...", "name": "appctle" }

}? yes

● create_project_api_v1_projects_post  id call_f26550z0
{
  "active": true,
  "description": "CLI app that works for devs and clients for free",
  "name": "appctle"
}
→ ok · 65ms · id=call_f26550z0
{
  "data": {
    "active": true,
    "created_at": "2024-04-23T12:43:34.391067",
    "description": "CLI app that works for devs and clients for free",
    "id": 1,
    "name": "appctle"
  },
  "status": 201
}

Project created successfully:
- ID: 1
- Name: appctle
- Description: CLI app that works for devs and clients for free
- Active: Yes

Terminal session: list projects as a table

list_projects after create, with formatted table

appctl[ollama]▶ List the projects now
● list_projects_api_v1_projects_get  id call_1dxqpvkz
{
  "limit": 10,
  "offset": 0
}
→ ok · 2ms · id=call_1dxqpvkz
{
  "data": [
    {
      "active": true,
      "created_at": "...",
      "description": "CLI app that works for devs and clients for free",
      "id": 1,
      "name": "appctle"
    }
  ],
  "status": 200
}

Here's the current project list:
┌────┬─────────┬─────────────────────────────────────────────┬────────┐
│ ID │ Name    │ Description                                 │ Active │
├────┼─────────┼─────────────────────────────────────────────┼────────┤
│ 1  │ appctle │ CLI app that works for devs and clients...  │ Yes    │
└────┴─────────┴─────────────────────────────────────────────┴────────┘

Web UI: same API, conversational formatting

With appctl serve, you can use the bundled console against the same tools. The model can follow formatting instructions (here, listing project labels with emojis, then replying with emojis only when asked).

Web chat with API status bar and emoji-heavy replies

License

MIT © Esubalew