Rising AI repos
Top 10 rising AI repos, week of Oct 19 (ET)
ET · 10 repos
Data: ET · As of · Published
Total stars are a snapshot at that time. Stars gained come from GitHub’s daily star data, whose day boundaries may not line up with ET midnight.
The short version
morluto/rea, an MCP server that lets agents inspect binaries and apps, gained +60,507 stars, about 2.7 times the runner-up. DeepSeek’s DeepInfer inference framework came second with +22,635, nearly all of its 23,093 total.
Four of the ten are agent skills, and two more are agent harnesses, from Google Research and DeepSeek.
-
morluto/rea
+60,507 stars gained. One MCP server that lets your agent inspect native binaries, JavaScript and Electron apps, .NET assemblies and websites, then explain how a feature works.
Why it matters for agent builders: It puts heavy local tools (Ghidra, IDA or Hopper) behind a single MCP, a pattern worth copying when you wire big desktop tools into an agent.
Source: morluto/rea on GitHub
-
deepseek-ai/DeepInfer
+22,635 stars gained. The inference framework behind DeepSeek’s production services, released for offline inference on DeepSeek-V4.1-Flash with structured outputs and deterministic sampling.
Why it matters for agent builders: Batch-invariant, deterministic runs could make agent evals easier to repeat; this release leaves out the KV cache and load balancing, so it is not a full serving stack.
Source: deepseek-ai/DeepInfer on GitHub
-
cloudflare/security-audit-skill
+16,231 stars gained. A coding-agent skill that runs a six-phase security audit with isolated hunter and verifier agents and writes findings against a JSON schema.
Why it matters for agent builders: A worked example of find-then-try-to-disprove, which may cut false positives in any agent that reports problems.
Source: cloudflare/security-audit-skill on GitHub
-
google-research/harness2
+11,711 stars gained. A research framework that tunes an agent’s harness (prompts, tools, memory, skills) per task without touching model weights, with OpenCode and Codex executors.
Why it matters for agent builders: If the reported benchmark gains hold up, tuning the harness may be a cheaper lever than switching models.
Source: google-research/harness2 on GitHub
-
ai-job-search-dev/ai-job-search
+11,700 stars gained. A job-application workflow built on Claude Code: slash commands that scrape postings, score fit and draft a tailored CV and cover letter.
Why it matters for agent builders: A clear template for shipping a multi-step personal workflow as Claude Code commands that people fork and fill in.
Source: ai-job-search-dev/ai-job-search on GitHub
-
DietrichGebert/ponytail
+9,382 stars gained. A skill that tells your coding agent to write only the code a task needs, reuse what the repo already has, and never cut validation, error handling or tests.
Why it matters for agent builders: Its own Claude Code benchmark claims less code and fewer tokens with tests kept; worth checking against your own tasks before you rely on it.
Source: DietrichGebert/ponytail on GitHub
-
stablyai/orca
+9,264 stars gained. A desktop and mobile workspace that runs Codex, Claude Code, OpenCode or Pi side by side, each in its own git worktree, tracked in one place.
Why it matters for agent builders: Fanning one prompt across several agents and keeping the best result is becoming a common workflow, and this shows one way to manage it.
Source: stablyai/orca on GitHub
-
deepseek-ai/deepseek-harness
+8,566 stars gained. DeepSeek’s open-source agent harness (dsh), built so that everything is a plugin; it is a developer preview with breaking changes expected.
Why it matters for agent builders: A model lab’s own harness shows which plugin points it thinks an agent needs; expect the APIs to move while it is in preview.
Source: deepseek-ai/deepseek-harness on GitHub
-
mattpocock/skills
+7,183 stars gained. A set of small, composable agent skills for everyday engineering work, installable as a plugin for Claude Code, Codex and GitHub Copilot.
Why it matters for agent builders: Short, editable skills you own are an alternative to process frameworks that take over the whole workflow.
Source: mattpocock/skills on GitHub
-
Graphify-Labs/graphify
+7,096 stars gained. A /graphify skill that turns a codebase, its docs, SQL schemas, configs and PDFs into a knowledge graph your assistant can query instead of grepping.
Why it matters for agent builders: Code is parsed locally with tree-sitter and every edge is labeled extracted or inferred, a different retrieval approach from vector search.
Source: Graphify-Labs/graphify on GitHub
What to watch next
- DeepInfer: nearly all of its 23,093 stars arrived in this window, so the next chart shows whether interest holds after launch.
- Agent skills: cloudflare/security-audit-skill, ponytail, mattpocock/skills and graphify all ship as skills; see which ones keep climbing.
- deepseek-harness is a developer preview with breaking changes expected, per its README.