The week in brief
Behind it, OpenAI Codex (+199) and Ollama (+196) followed. This first comparison spans 1.5 days; from next week each report compares a full seven days.
On Hacker News, “RIP, vector database” drew the most attention in this space (401 points, 116 comments). Among listed tools, Cursor's site was linked from 37 Hacker News stories over the past 90 days, more than any other tool here.
In research, “From Evidence to Action: How Tool-Using Agents Fail” collected the most upvotes on Hugging Face Daily Papers this week (30). Among new GitHub repositories created in the past 7 days, GTKottman/mortiflix-oss leads with 285 stars.
Biggest GitHub star gainers
GitHub star counts for the 48 open-source tools we track, compared between Oct 5, 2026 and Oct 7, 2026 (37 hours apart).
| Tool | Stars before | Stars after | Gain | Change |
|---|---|---|---|---|
| Firecrawl | 188,919 | 189,368 | +449 | 0.24% |
| OpenAI Codex | 127,950 | 128,149 | +199 | 0.16% |
| Ollama | 182,263 | 182,459 | +196 | 0.11% |
| Claude Code | 149,524 | 149,715 | +191 | 0.13% |
| Mem0 | 66,622 | 66,763 | +141 | 0.21% |
| LiteLLM | 60,169 | 60,283 | +114 | 0.19% |
| Dify | 157,898 | 158,009 | +111 | 0.07% |
| vLLM | 93,225 | 93,327 | +102 | 0.11% |
Most discussed on Hacker News
The highest-scoring Hacker News stories in this space during the week.
| Story | Points | Comments | Discussion |
|---|---|---|---|
| RIP, vector database | 401 | 116 | thread |
| Figma restricts MCP access to whitelisted clients, excluding Pi | 188 | 106 | thread |
| Show HN: Pi pod – Run your pi coding agent in sandboxes on your own server | 121 | 47 | thread |
| The Four Horsemen of Agentic Coding | 114 | 89 | thread |
| South Korea says AI agents appear to have been used to hack the country's banks | 88 | 25 | thread |
| Identity Management for Agentic AI [pdf] (2025) | 82 | 28 | thread |
Listed tools with the most Hacker News links
Hacker News stories linking to each tool's own website over the past 90 days.
Notable new papers
The most-upvoted papers on Hugging Face Daily Papers and the newest relevant arXiv preprints from this week.
- From Evidence to Action: How Tool-Using Agents Fail
- AutoSciBench: Autonomous Benchmark Generation for Evaluating Scientific Agents
- MiniCorp: The Last Mile of the AI Agent Firm
- NeMo-DCR: Bit-Exact Delta-Compressed Refit for Scalable Agentic RL at Trillion-Parameter Scale
- DAEDALUS: Bootstrapping Agent Memory from Self-Generated Tasks
- Agent in a Bottle: Can LLM Agents Turn Their Capabilities Into Cheap, Scalable Artifacts?
- AdvSim2Real : Training Web Agents Against Adaptive Prompt Injection in a Web World Model
- Does an Agent's History Tell You When Compaction Will Hurt? A Modest, Bounded Effect on the TRACE Paired-Replay Corpus
- WorldSolver: Can LLM Agents Simulate the Physical Dynamics via Solver Generation?
- nanoMuse: An Open-Source Personal Agent for Every Device You Own
New GitHub repos in AI agents
Repositories created in the 7 days before Oct 07, 2026, ranked by stars.
| Repository | Stars | Language | Description |
|---|---|---|---|
| GTKottman/mortiflix-oss | 285 | JavaScript | A motion design studio on your own machine: Claude makes the video step by step, you approve every stage. Bring your own Claude Code or API key. |
| mariagorskikh/open-instinct | 231 | TypeScript | Open Instinct: an open-source Instinct. A personal agent you text on iMessage, with its own computer and a trusted network of other agents. Built on Pi, Inkbox, Maritime and Composio. |
| uzairansaruzi/p3-stack | 136 | Shell | pstack reworked for T3 Code: delegated subagents, worktree threads, PR watching, scheduled runs. |
| xop01/ai_goodpractice | 136 | full guide to help you integrate AI in your workflow, for both beginners and experienced users | |
| hamzafer/claude-code-mods | 135 | TypeScript | Claude Code mods: plugins built on hooks that add live lines above the prompt, guards, panes and games. Context bar, usage meter, Codex review watch, Markdown preview, Spotify now playing and more. |
| nickvasilescu/bops | 131 | TypeScript | Bops: a team of AI bots that run your business ops, each with its own computer, email and phone number. bops.bot |
| hyj-STAR/voiceshell-muse-bridge | 126 | Python | VoiceShell（声壳） · Experimental Muse bridge: ESP32 pairing, Linux text round trips, real walkthroughs and compatibility notes. |
| openpactprotocol/openpactprotocol | 85 | TypeScript | PACT is an open protocol for personal agents to interact with businesses’ agents under explicit, verifiable user permissions. |
How we compile this report
Every figure comes from a public source: star counts from the GitHub API, story scores and links from Hacker News (via Algolia), papers from arXiv and Hugging Face Daily Papers, and new repositories from GitHub search. Hacker News link counts include stories whose link points to the tool's own domain. Listing dates are the day a tool joined Drydock. Numbers are shown as recorded; if a figure is unavailable, it is left out rather than estimated.