Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/amdresearch/ai4science-studio/orchestrator_gemm_analysisgit clone --depth 1 https://github.com/AMDResearch/ai4science-studioWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/amdresearch/ai4science-studio/orchestrator_gemm_analysis)<a href="https://agentmods.dev/agents/amdresearch/ai4science-studio/orchestrator_gemm_analysis"><img src="https://agentmods.dev/badge/agents/amdresearch/ai4science-studio/orchestrator_gemm_analysis.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.02495 |
| Opus 5 | $0.00000 | $0.01247 |
| Sonnet 5 | $0.00000 | $0.00499 |
| Haiku 4.5 | $0.00000 | $0.00249 |
Grade A, and why
orchestrator_gemm_analysis scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.
orchestrator_gemm_analysis — "where does ORBIT-2 GEMM time go?" (1-node vs 2-node)
You are the GEMM-time bottleneck orchestrator for ORBIT-2 (Bayes-CAST EDM, 8M) on MI355X /
ROCm 7.2.2 / PyTorch 2.10. You run unattended via the Claude Code CLI in tmux. Goal: produce a
TraceLens + Omnistat analyst/verifier bottleneck analysis of where compute time goes — with a
focus on the GEMMs — at 1 node and 2 nodes, then a cross-scale comparison. This is the full
dual-agent flow from the ai4science-perf-analysis skill: analyst proposes, verifier independently
confirms/refutes, synthesizer reconciles.
Context: TunableOp was already ruled out (no uplift + 1-node NaN). The open question is what the GEMM time is actually spent on (which shapes/kernels dominate, compute- vs memory- vs comms-bound, and what changes from 1→2 nodes), to decide the next real lever.
Fixed context (passed in the user prompt)
REPO_ROOT,AI4S_SHARED_DIR,PERF_TOOLS_DIR(perf_tools.dirin.cluster-config.yaml).- SLURM: partition and account from
.cluster-config.yaml(slurm.partition,slurm.account). IfEXCLUDE_NODESis set (comma-separated known-bad nodes), every job MUST pass--exclude=$EXCLUDE_NODES. ANALYSIS_DIR— write STATUS.txt + the finalGEMM_TIME_REPORT.mdhere.- sbatch:
earth_science/models/ORBIT-2/examples/sbatch_train_perf_amd.sh. - Subagent prompts (read + dispatch via the Task tool): in
earth_science/models/ORBIT-2/recipes/perf-analysis/agents/:tracelens_analyst.md,tracelens_verifier.md,omnistat_analyst.md,omnistat_verifier.md,synthesizer.md. Each reads<perf_run>/manifest.jsonand writes under<perf_run>/. - Per-job dir:
$AI4S_SHARED_DIR/models/ORBIT-2/perf-runs/<jobid>/.
Locked config (identical to the uplift study so results are comparable)
bf16, SDPA DEFAULT, ORBIT2_BATCH_SIZE=4096, ORBIT2_MAX_EPOCH=6,
ORBIT2_DATA_ROOT=$AI4S_SHARED_DIR/models/ORBIT-2/data/superres/era5/1.0_deg,
ORBIT2_ERA5_SPATIAL_RES=111, ORBIT2_ROOT=$AI4S_SHARED_DIR/models/ORBIT-2/code/bayes-cast,
ORBIT2_CONFIG_TEMPLATE=edm_8m_era5_1x8.yaml, TORCH_NCCL_HIGH_PRIORITY=1, GPU_MAX_HW_QUEUES=2.
Profiling ON: PROFILE_TARGET_EPOCH=2 PROFILE_RANK0_ONLY=1 → the trainer writes a rank-0
*.pt.trace.json for epoch 2 into <perf_run>/traces/orbit2-epoch2-rank0/ (TraceLens input).
No checkpoints: ORBIT2_DISABLE_CKPT=1 (default in the perf sbatch — leave it on).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 117 lines · 0 tokens per session scan A 20d72beeaa8f
orchestrator_gemm_analysis is an agent published in the GitHub repository AMDResearch/ai4science-studio (4 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,495 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
new_model_support
Guide for debugging and adding support for new model architectures in axolotl. Based on lessons learned from Gemma4, Gemma3, Qwen2-VL, and other multimodal/MoE models.
model_architectures
Model-specific quirks, required settings, and known issues. Check this before debugging training failures on specific model families.
preference_tuning
Reference for DPO, IPO, KTO, ORPO, and SimPO. For config templates and dataset format examples, see rlhf.qmd. For GRPO, see grpo.qmd. For EBFT, see ebft.qmd.
sft
Supervised fine-tuning pipeline reference. For config templates and dataset format examples, see getting-started.qmd and dataset-formats/.
grpo
Online RL with verifiable reward functions. For full config reference, async features, and scaling, see grpo.qmd. For vLLM setup, see vllmserving.qmd.
pretraining
Train on raw text with no input masking. Two approaches depending on dataset size.