Add Uint Support
pytorch/pytorch
Add unsigned integer (uint) type support to PyTorch operators by updating ATDISPATCH macros.
Launch an autonomous THINK→EXECUTE→REFLECT experiment loop on a GPU project
Install Auto Experiment
$ npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install Xiangyue-Zhang/auto-deep-researcher-24x7 auto-experiment --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/auto-experiment .claude/skills/auto-experiment && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "auto-experiment" agent skill from https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experiment into .claude/skills/auto-experiment/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "auto-experiment", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experimentType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install Xiangyue-Zhang/auto-deep-researcher-24x7 auto-experiment --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/auto-experiment .agents/skills/auto-experiment && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "auto-experiment" agent skill from https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experiment into .agents/skills/auto-experiment/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "auto-experiment", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install Xiangyue-Zhang/auto-deep-researcher-24x7 auto-experiment --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/auto-experiment .cursor/skills/auto-experiment && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "auto-experiment" agent skill from https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experiment into .cursor/skills/auto-experiment/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "auto-experiment", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git --path skills/auto-experiment--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install Xiangyue-Zhang/auto-deep-researcher-24x7 auto-experiment --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/auto-experiment .gemini/skills/auto-experiment && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "auto-experiment" agent skill from https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experiment into .gemini/skills/auto-experiment/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "auto-experiment", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install Xiangyue-Zhang/auto-deep-researcher-24x7 auto-experimentInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/auto-experiment .github/skills/auto-experiment && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "auto-experiment" agent skill from https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experiment into .github/skills/auto-experiment/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "auto-experiment", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install Xiangyue-Zhang/auto-deep-researcher-24x7 auto-experiment --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/auto-experiment .opencode/skills/auto-experiment && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "auto-experiment" agent skill from https://github.com/Xiangyue-Zhang/auto-deep-researcher-24x7/tree/main/skills/auto-experiment into .opencode/skills/auto-experiment/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "auto-experiment", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
auto-experiment- GitHub stars
- 1.3k
- Token cost
- ~1.4k tokens
- SKILL.md length
- 428 words
- Files
- 2
- Skills in repo
- 8
- Repo updated
- First seen
- Licence
- Apache-2.0
At a glance
Launch an autonomous THINK→EXECUTE→REFLECT experiment loop on a GPU project
- Works in 3 steps: THINK → EXECUTE → REFLECT
- Tasks that involve Deep learning
- SKILL.md covers What This Does, Usage, Prerequisites and Workflow Details, plus 3 more sections
- Calls claude
What it does
Auto Experiment is an agent skill from Xiangyue-Zhang/auto-deep-researcher-24x7. Launch an autonomous THINK→EXECUTE→REFLECT experiment loop on a GPU project
Its SKILL.md is about 1.4k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).
It sits in AI & LLM Engineering, covering Deep learning. The repository describes itself as: 🔥 An autonomous AI agent that runs your deep learning experiments 24/7 while you sleep. Zero-cost monitoring, Leader-Worker architecture, constant-size memory. The licence is Apache-2.0.
When your agent uses it
- Tasks that involve Deep learning
Example prompts
- “/auto-experiment”
Workflow steps
3 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit dbf3df8. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
claude
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Auto Experiment loads about 1.4k tokens when it runs. Until then it costs about 23 tokens; SKILL.md has 428 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from Xiangyue-Zhang/auto-deep-researcher-24x7 at commit dbf3df8, republished under its Apache-2.0 licence (© Xiangyue-Zhang). 428 words, ~1,396 tokens.
Download SKILL.mdSave it as
.claude/skills/auto-experiment/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.- name
- auto-experiment
- description
- Launch an autonomous THINK→EXECUTE→REFLECT experiment loop on a GPU project
auto-experiment
Launch an autonomous experiment agent that runs your deep learning experiments 24/7.
What This Does
This skill starts a THINK → EXECUTE → REFLECT loop that:
- Reads your
PROJECT_BRIEF.mdto understand the research goal - Analyzes previous results in
MEMORY_LOG.md - Plans the next experiment (hypothesis + success criteria)
- Implements code changes and runs a mandatory dry-run
- Launches GPU training via
nohup(tracks PID) - Monitors at zero LLM cost (only
kill -0 PID+tail log+nvidia-smi) - Wakes up when training finishes to analyze results
- Updates memory and decides: iterate, pivot, or report
- Repeats
Usage
Claude Code: /auto-experiment
Claude Code: /auto-experiment --project /path/to/my_project --gpu 0
Claude Code: /auto-experiment --project . --max-cycles 5
Codex: $auto-experimentPrerequisites
The project directory must contain:
PROJECT_BRIEF.md (required)
A frozen reference describing your research goal. Example:
# Goal
Train a ViT-B/16 on ImageNet to reach 78%+ top-1 accuracy.
# Codebase
- Training: train.py
- Config: configs/vit_base.yaml
- Data: /data/imagenet/
# Constraints
- GPU 0-3 available (use DDP)
- Max 90 epochs per run
- Report val accuracy after each run
# Current Best
- ResNet-50 baseline: 76.1%config.yaml (optional)
Override default agent settings:
agent:
provider: "anthropic" # or "openai" / "claude_cli" / "codex_cli"
model: "claude-sonnet-4-6"
base_url: "" # optional compatible endpoint override
api_key_env: "" # optional custom key env var
auth_token_env: "" # optional custom bearer token env var
max_cycles: -1 # -1 = unlimited
max_steps_per_cycle: 3 # max sub-agent dispatches per cycle
cooldown_interval: 300 # 5 min smart polling
memory:
brief_max_chars: 3000
log_max_chars: 2000
monitor:
poll_interval: 900 # check every 15 min during training
zero_llm: true
experiment:
mandatory_dry_run: trueIf the user wants a compatible API endpoint instead of the official Anthropic
or OpenAI API, keep the same provider values and set base_url plus a custom
api_key_env. Do not invent provider names like qwen or glm.
Optional remote execution over SSH:
execution:
mode: "ssh"
ssh_host: "user@server"
remote_workspace: "/home/user/my_project/workspace"
remote_python: "python3"In SSH mode, the controller state stays local (PROJECT_BRIEF.md,
workspace/MEMORY_LOG.md, workspace/HUMAN_DIRECTIVE.md, state.json),
while code edits, shell commands, training, log tailing, PID checks, and GPU
queries run on the configured remote host.
Workflow Details
Phase 1: THINK
- Read
PROJECT_BRIEF.md(frozen, max 3000 chars) - Read
MEMORY_LOG.md(rolling, auto-compacted) - Check for
HUMAN_DIRECTIVE.md(highest priority, auto-archived after reading) - Analyze: What's the current best? What hasn't been tried? What's most promising?
- Output: experiment plan with hypothesis and success criteria
Show full SKILL.md (191 more words)Show less
Phase 2: EXECUTE
- Dispatch to Code Agent (5 tools:
run_shell,launch_experiment,write_file,read_file,list_files) - Code Agent implements changes
- Mandatory dry-run (2-step verify, abort if fails)
- Launch training via
nohup, capture PID - Enter zero-cost monitoring loop:
- backend PID check — is process alive?
- backend
nvidia-smi— GPU utilization - backend
tail -50 logfile— latest training output - Zero LLM API calls during this phase
Phase 3: REFLECT
- Parse training logs for metrics (loss, accuracy, FGD, FID, etc.)
- Compare against previous best
- Log milestone if improved (auto-compacted at 1200 chars)
- Log decision (rolling last 15 entries)
- Decide: try another config / pivot direction / generate report
Human Override (anytime)
# Drop a directive file — agent reads it next cycle with highest priority
echo "Try learning rate 1e-5 with cosine schedule" > workspace/HUMAN_DIRECTIVE.mdMemory System
Two-Tier, constant size (~5K chars / ~1500 tokens), no matter how long the agent runs:
| Tier | File | Content | Cap |
|---|---|---|---|
| 1 | PROJECT_BRIEF.md | Frozen project reference | 3,000 chars |
| 2 | MEMORY_LOG.md | Key Results + Recent Decisions | 2,000 chars |
Auto-compaction rules:
- Key Results: oldest dropped when section > 1,200 chars
- Recent Decisions: only last 15 entries kept
- Total log hard-capped at 2,000 chars
Cost
| Phase | Duration | LLM Cost |
|---|---|---|
| THINK | 5-10 min | ~$0.05 |
| EXECUTE (training) | hours/days | $0.00 |
| REFLECT | 5-10 min | ~$0.03 |
| 24h cycle total | ~$0.08 |
Example Output
After a few cycles, your workspace/MEMORY_LOG.md will look like:
# Memory Log
## Key Results
[04-07 14:30] Exp001: ResNet-50 baseline, lr=0.1, acc=76.1%
[04-07 22:15] Exp002: ViT-B/16, lr=1e-3, acc=74.8% (underperforming, lr too high)
[04-08 06:00] Exp003: ViT-B/16, lr=3e-4 + cosine, acc=77.9% (new best!)
[04-08 14:45] Exp004: ViT-B/16, lr=3e-4 + cosine + mixup, acc=78.3% (target reached!)
## Recent Decisions
[04-07 14:30] Start with ResNet-50 baseline to establish reference
[04-07 22:15] ViT lr=1e-3 too high, try 3e-4 next
[04-08 06:00] Cosine schedule helped significantly, try adding regularization
[04-08 14:45] Target reached! Generate final report.© Xiangyue-Zhang, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 1 other file in skills/auto-experiment of Xiangyue-Zhang/auto-deep-researcher-24x7.
- SKILL.md
- agents/openai.yaml
Open the folder on GitHubat commit dbf3df8
Compare with similar skills
Auto Experiment next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Auto Experiment this skillXiangyue-Zhang/auto-deep-researcher-24x7 | 1.3k | — | ~1.4k | Automated safety check: Pass | Apache-2.0 | |
| Add Uint Supportpytorch/pytorch | 104k | 2 repos | ~2.3k | Automated safety check: Pass | Custom licence | |
| Segment Anything Model GuideOrchestra-Research/AI-Research-SKILLs | 13k | 8 repos | ~3.3k | Automated safety check: Pass | MIT | |
| Add Oponnx/onnx | 22k | — | ~1.2k | Automated safety check: Pass | Apache-2.0 | |
| CLIP Image-Text MatchingOrchestra-Research/AI-Research-SKILLs | 13k | 7 repos | ~1.7k | Automated safety check: Pass | MIT | |
| Add Function Bodyonnx/onnx | 22k | — | ~1.1k | Automated safety check: Pass | Apache-2.0 |
Similar skills
Segment Anything Model Guide
Orchestra-Research/AI-Research-SKILLs
Guide to using Meta's Segment Anything Model for zero-shot image segmentation with point, box or mask prompts, or automatic mask generation.
Add Op
onnx/onnx
Add a new ONNX operator or update an existing operator to a new opset version.
CLIP Image-Text Matching
Orchestra-Research/AI-Research-SKILLs
Explains OpenAI's CLIP model for zero-shot image classification, image-text similarity, semantic image search and content moderation, with install steps and code patterns.
Add Function Body
onnx/onnx
Add a function body definition to an ONNX operator, defining how it decomposes into simpler ops.
Paddle Design Distributed
PaddlePaddle/Paddle
A skill your agent uses when working with Paddle's distributed training system: understanding parallelism strategies (DP, ZeRO, TP, PP, SP), semi-automatic parallel with ProcessMesh + shardtensor…
More from Xiangyue-Zhang/auto-deep-researcher-24x7
All 8 skills in this repoConf Search
Xiangyue-Zhang/auto-deep-researcher-24x7
Search papers from top AI/ML conferences. An agent skill from Xiangyue-Zhang/auto-deep-researcher-24x7.
Daily Papers
Xiangyue-Zhang/auto-deep-researcher-24x7
Daily arXiv paper recommendations with automatic deduplication
Experiment Status
Xiangyue-Zhang/auto-deep-researcher-24x7
Check status of running autonomous experiment loops. An agent skill from Xiangyue-Zhang/auto-deep-researcher-24x7.
GPU Monitor
Xiangyue-Zhang/auto-deep-researcher-24x7
Check GPU status, running experiments, and available resources
Obsidian Sync
Xiangyue-Zhang/auto-deep-researcher-24x7
Refresh Obsidian dashboard and daily notes from current experiment state
Paper Analyze
Xiangyue-Zhang/auto-deep-researcher-24x7
Deep analysis of a single paper with figure extraction from arXiv source
Related
Categories
Questions about Auto Experiment
What does Auto Experiment do?
Launch an autonomous THINK→EXECUTE→REFLECT experiment loop on a GPU project. Auto Experiment is an agent skill from Xiangyue-Zhang/auto-deep-researcher-24x7.
When should I use Auto Experiment?
Auto Experiment fits situations like: tasks that involve Deep learning.
How do I install Auto Experiment in Claude Code?
Run `npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a claude-code`. Or copy the skill folder (skills/auto-experiment in Xiangyue-Zhang/auto-deep-researcher-24x7) into .claude/skills/auto-experiment in your project. Claude Code loads it when a task matches its description.
How do I install Auto Experiment in Codex?
Run `npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a codex`. Or copy the skill folder (skills/auto-experiment in Xiangyue-Zhang/auto-deep-researcher-24x7) into .agents/skills/auto-experiment in your project. Codex loads it when a task matches its description.
Can I use Auto Experiment in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Xiangyue-Zhang/auto-deep-researcher-24x7 --skill auto-experiment -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/auto-experiment, .gemini/skills/auto-experiment, .github/skills/auto-experiment and .opencode/skills/auto-experiment in your project.
What does Auto Experiment need to run?
Going by SKILL.md and its folder, Auto Experiment needs the command-line tools its instructions call (claude).
Does Auto Experiment access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Auto Experiment safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Auto Experiment use?
Auto Experiment is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Auto Experiment use?
About 1.4k tokens (SKILL.md is roughly 5.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Auto Experiment?
Skills that share tags, products or a category with Auto Experiment: Add Uint Support (pytorch/pytorch, 104k stars), Segment Anything Model Guide (Orchestra-Research/AI-Research-SKILLs, 13k stars), Add Op (onnx/onnx, 22k stars) and CLIP Image-Text Matching (Orchestra-Research/AI-Research-SKILLs, 13k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Auto Experiment?
Xiangyue-Zhang (a GitHub user) maintains it in Xiangyue-Zhang/auto-deep-researcher-24x7, which has 1,296 GitHub stars. The repository holds 8 skills in this directory. The repository was last updated on June 3, 2026.
Source: Xiangyue-Zhang/auto-deep-researcher-24x7 on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.