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 skills/duonginspace/claude-code-databricks-ml/train-localnpx skills add duonginspace/claude-code-databricks-ml --skill train-localgit clone --depth 1 https://github.com/duonginspace/claude-code-databricks-mlWrote 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/skills/duonginspace/claude-code-databricks-ml/train-local)<a href="https://agentmods.dev/skills/duonginspace/claude-code-databricks-ml/train-local"><img src="https://agentmods.dev/badge/skills/duonginspace/claude-code-databricks-ml/train-local.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 | $0.00056 | $0.00310 |
| Opus 5 | $0.00028 | $0.00155 |
| Sonnet 5 | $0.00011 | $0.00062 |
| Haiku 4.5 | $0.00006 | $0.00031 |
Grade A, and why
train-local 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 4d 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.
What it actually says
Local training task
$ARGUMENTS
Steps
- Read
CLAUDE.mdfor the project structure and current experiment config. - If not specified, default to: 2 epochs, 10% of training data, CPU/MPS device, batch size 16.
- Add
--dry-runor--fast-dev-runflag if the training script supports it. - Run:
uv run python scripts/train.py <args> 2>&1 | tee mlflow_results/local_run.log - Watch for: import errors, shape mismatches, CUDA/MPS device errors, NaN losses, OOM errors.
- If the run succeeds, report: final train loss, validation metric, time per epoch.
- If it fails, diagnose the error and propose a fix before suggesting a Databricks run.
Output
Tell the user whether the code is ready to submit to Databricks, or what to fix first.
Always suggest using /run-on-databricks after a successful local smoke test.
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.
- 4d ago First seen · 24 lines · 56 tokens per session scan A 09930322af35
train-local is a skill published in the GitHub repository duonginspace/claude-code-databricks-ml (5 stars, last pushed 5mo ago), licensed MIT. It adds 56 tokens to every session and 310 once invoked, about $0.0003 per session on Opus 5. 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 skills, from other repositories
exploratory-data-analysis
Perform bounded, local exploratory analysis of explicitly supported scientific files. Use for redacted CSV/TSV/JSON profiles; optional NumPy, HDF5, FASTA/FASTQ, and basic image metadata inspection; missingness/leakage audits; outlier and transformation sensitivity; and rigorous EDA report scaffolds. Other domain…
evaluating-with-leakage-gates
Evaluate an OpenMed de-identification or clinical NER model against the leakage-first release gates G1a through G8, which gate releases on residual PHI leakage rather than on F1. Use when the user wants to run the OpenMed eval harness on a synthetic golden set, decide whether a de-id model is RELEASABLE or…
mapping-to-snomed
Maps clinical concept spans extracted by OpenMed to SNOMED CT concepts through a USER-SUPPLIED terminology server (the user's own Ontoserver, Snowstorm, or UMLS/UTS), never a bundled vocabulary. Use when the user wants to code findings, disorders, procedures, body structures, or substances to SNOMED CT, run an ECL…
mixed-precision
Use FP16/BF16 mixed precision to accelerate training and reduce memory. Use when optimizing GPU performance.
tooluniverse-drug-research
Comprehensive drug profiling — mechanism, primary/secondary targets, drug interactions, clinical-trial status, adverse events (FAERS), pharmacogenomics, and approval history. Use for full drug investigation reports, 'tell me about drug X' queries, and assembling drug profiles for clinicians, researchers, or regulatory…
audit-scientific-figure
Review and score an existing scientific illustration in visible draw.io, Microsoft PowerPoint, or WPS Presentation without hiding defects through flattening. Use for reference-fidelity checks, layout cleanup, connector review, text-fit checks, deep editability and raster-atomicity inspection, local-region gates, or…