synthetic-sciences/openscience is an AI workbench that carries out scientific research by reading papers, forming hypotheses, writing and running code, conducting experiments, analyzing results, and preparing reports. Researchers use it for work in machine learning, biology, physics, and chemistry with remote or local models. Catalogue add-ons extend its scientific workflows through skills and instructions.
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 skills add synthetic-sciences/openscience --skill autoresearchgit 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/autoresearch)<a href="https://agentmods.dev/skills/synthetic-sciences/openscience/autoresearch"><img src="https://agentmods.dev/badge/skills/synthetic-sciences/openscience/autoresearch/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/synthetic-sciences/openscience/autoresearch"><img src="https://agentmods.dev/badge/skills/synthetic-sciences/openscience/autoresearch.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00110 | $0.01647 |
| Opus 5.5 | $0.00044 | $0.00659 |
| Sonnet 5.5 | $0.00022 | $0.00329 |
| Haiku 4.5 | $0.00011 | $0.00165 |
Grade A, and why
autoresearch 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 20d 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 — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Autoresearch
A study turns "make this metric better" into a loop that runs for hours or days without losing the thread. You own the science: which ideas to try, how to implement them, what a result means. OpenScience owns the clock: it tracks every metric the script logs, notices when a run ends, kills runs that break the study's criteria, keeps the budget, renders the ledger, and wakes this session with a "Study update" whenever there is news.
Before the study
- Agree the objective with the user: one metric, its direction, the budget, the compute target, how many runs may be live at once, and the kill criteria. The budget is a spend decision and belongs to this study alone: never carry one over from an earlier study or an earlier instruction in the session. If the request names no budget, ask once, with a recommendation (a time budget such as 2 hours with a per-run kill rule, or a target value), even under an autonomous setting. Everything else, infer or default.
- Read the code and data first. A study needs a training or analysis script the harness can run repeatedly with different configuration, and a fixed evaluation that computes the metric. The evaluation does not change once the baseline has run.
- Create the study with
study create. Propose the baseline with priority 1000 and at least three first ideas withstudy propose, of different kinds (a different component, objective, data treatment or search strategy each), not three magnitudes of one knob. An idea has a title, what it changes, why it should help, an expected improvement in metric units times your confidence, and the configuration it needs; a configuration already in the study is rejected. The hypotheses skill is the tool for turning a vague direction into ideas worth queueing. - If the review gate is on (the default), get a read-only critique of the training and
evaluation code before the baseline runs: load the peer-review skill and read it
yourself, or delegate one read to an
exploreworker (Task tool,subagent_type: "explore") with the file paths and the angle (leakage, metric definition, split hygiene). Fix anything marked blocking. A baseline built on a leaking split or a misspelled metric wastes every run after it.
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.
- 20d ago Changed · +1 lines 730a7f582683
- 1mo ago Changed · +4 lines f9220094ebd3
- 1mo ago Changed · +5 lines dd482562ce14
- 1mo ago Changed · +2 lines 9806468c6c9a
- 1mo ago First seen · 98 lines · 110 tokens per session scan A 4682bbb0d6da
autoresearch is a skill published in the GitHub repository synthetic-sciences/openscience (4,025 stars, last pushed today), licensed Apache-2.0. It adds 110 tokens to every session and 1,647 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-09-14.
Other skills, from other repositories
knowledge-extension-ecosystem-lsp-code-intelligence
A codebase guide to LSP, the standard protocol behind editor features such as symbol queries, file diagnostics, and language-aware code information.
github-explorer
Explore and summarize popular GitHub projects. Use when the user wants to discover trending GitHub repositories, analyze project details, or get summaries of popular projects.
google-workspace-setup
One-time setup for gws: install, OAuth, scopes, auto-approve.
google-sheets
Google Sheets via gws: read/write cells, append rows, structured batch edits.
attachments
Move bytes between Gini upload space, external URLs, and workspace files. Used by every attachment / file-upload / file-download flow regardless of the target system (Linear, GitHub, S3, Notion, etc.).
google-docs
Google Docs via gws: read, append text, structured batch edits.