No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by aisa-group · Codex Skill · ★ 545
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is PostTrainBench safe to install? View the security audit →
PostTrainBench: Can LLM Agents Automate LLM Post-Training? We introduce PostTrainBench, a benchmark that measures the ability of CLI agents to post-train pre-trained large language models (LLMs). In PostTrainBench, the agent's task is to improve the performance of a base LLM on a given benchmark. The agent is given access to an evaluation script and 10 hours on an H100 GPU. Performance is measured by the benchmark score of the post-trained LLM. This setup naturally evaluates an agent's ability to conduct AI R&D. [!IMPORTANT] Harbor support coming soon! This repository currently targets our internal HPC cluster (HTCondor). We are adding Harbor support to make it straightforward to run on rented hardware (e.g., cloud GPUs). See our PR. Leaderboard Scores are weighted averages across 7 benchmarks and 4 models (Qwen3-1.7B, Qwen3-4B, SmolLM3-3B, and Gemma-3-4B). Agents with multiple runs show averaged results. 25.5
| Stars | 545 |
| Forks | 63 |
| Language | Python |
| Category | Codex Skill |
| License | MIT |
| Quality Score | 67.4302057640871/100 |
| Open Issues | 20 |
| Last Updated | 2026-09-02 |
| Created | 2025-11-28 |
| Platforms | claude-code, cli, codex, gemini, python |
| Est. Tokens | ~16k |
These tools work well together with PostTrainBench for enhanced workflows:
Looking for a PostTrainBench alternative? If you're comparing PostTrainBench with other codex skill tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Agent orchestration & security template featuring MCP tool building, agent2agent workflows, mechanistic interp
The control plane for AI coding agents.
A curated list of awesome LLM and AI Agent Skills, resources and tools for customising AI Agent workflows - th
Engineering decisions engine that know when they're stale. Frame, compare, decide — with evidence decay and p
Coordinate your coding agents like a group chat — read receipts, delivery tracking, and remote ops from your p
Battle-tested Claude Code, OpenAI Codex, Cursor configs, plugins, hooks and agents with Kimi, MiniMax and GLM
Explore other popular codex skill tools:
PostTrainBench is Measuring how well CLI agents like Claude Code or Codex CLI can post-train base LLMs on a single H100 GPU in 10 hours. It is categorized as a Codex Skill with 545 GitHub stars.
PostTrainBench is primarily written in Python. It covers topics such as ai-research-automation, ai-safety, claude-code.
You can find installation instructions and usage details in the PostTrainBench GitHub repository at github.com/aisa-group/PostTrainBench. The project has 545 stars and 63 forks, indicating an active community.
PostTrainBench is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to PostTrainBench on Agent Skills Hub include template-repo, ai-devkit, awesome-llm-skills. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: