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/synthetic-sciences/openscience/tensorpoolnpx skills add synthetic-sciences/openscience --skill tensorpoolgit clone --depth 1 https://github.com/synthetic-sciences/openscienceWrote 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/synthetic-sciences/openscience/tensorpool)<a href="https://agentmods.dev/skills/synthetic-sciences/openscience/tensorpool"><img src="https://agentmods.dev/badge/skills/synthetic-sciences/openscience/tensorpool.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 | $0.00034 | $0.00861 |
| Opus 5 | $0.00017 | $0.00430 |
| Sonnet 5 | $0.00007 | $0.00172 |
| Haiku 4.5 | $0.00003 | $0.00086 |
Grade A, and why
tensorpool-gpu-cloud 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 yesterday.
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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TensorPool GPU Cloud
Use this skill for TensorPool cluster, job, and storage workflows. Treat provider availability, instance names, quotas, and prices as live data: inspect them at execution time and never rely on values copied into a skill.
OpenScience credential boundary
- A key saved in Customize > Compute > TensorPool is encrypted control-plane data. It is not exported into Bash, Task, notebooks, kernels, plugins, or MCP servers.
- Use
provider_computewithaccount,list_resources,resource_status,list_jobs, orjob_statusfor live reads. OpenScience selects the exact reviewedtpargv, admits the key only to that isolated child, and markslast_usedonly after success. - Test connection uses the same boundary for exactly
tp --no-input me. - The saved key is admitted only when Test connection approves a root/admin-managed, non-writable
tpexecutable. A normal user-owned pip, pipx, or Homebrew install is credential-only and cannot be used byprovider_compute; use an administrator-managed install or wait for a managed/native adapter. - Never print, echo, persist, or ask the user to paste a saved key into chat.
- A generic agent shell can use
tponly when the user has independently authenticated that shell. The saved credential cannot create, submit, edit, attach, destroy, or otherwise mutate resources; do not ask the user to weaken this boundary.
Current CLI contract
For an independently authenticated user shell, install and inspect the CLI rather than guessing its version. This does not make a user-owned install eligible for the saved-key broker:
python -m pip install tensorpool
tp --help
tp --no-input me
The first-party CLI reads TENSORPOOL_KEY. The read-only identity check is tp --no-input me.
The broker maps its read operations to these current official commands:
tp cluster list
tp job list
tp storage list
storage list is not yet exposed by the broker. Run it only in an independently authenticated user shell.
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.
- yesterday First seen · 70 lines · 34 tokens per session scan A 0e4b7b91321a
tensorpool-gpu-cloud is a skill published in the GitHub repository synthetic-sciences/openscience (3,432 stars, last pushed yesterday), licensed Apache-2.0. It adds 34 tokens to every session and 861 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
architecture-diagram
Generate dark-themed SVG architecture diagrams as self-contained HTML.
skypilot-multi-cloud-orchestration
Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers.
lambda-labs-gpu-cloud
Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances with simple SSH access, persistent filesystems, or high-performance multi-node clusters for large-scale training.
cloudflare-router
Use when manage Cloudflare DNS, CDN, and security rules via API. Use when configuring domains, SSL, WAF, or edge caching.
latchbio-integration
Build, register, debug, and operate bioinformatics workflows on Latch using the Python SDK, CLI, Latch Data and Registry, Nextflow, Snakemake, programmatic execution, and Latch MCP. Use when authoring or deploying Latch workflows, configuring resources or interfaces, moving data, integrating Registry, or launching and…
modal
Modal is a serverless cloud platform for running Python on demand, including on-demand GPUs. Use when deploying or serving AI/ML models, running GPU-accelerated workloads (training, fine-tuning, inference), serving web endpoints, scheduling batch jobs, or scaling Python code to cloud containers with the Modal SDK.