research-strategy-and-project-design
Skill2 starsChanged 31 days agoMITView on GitHub
Choose scientific questions, hypotheses, experimental programs, or project continuation. Exclude analysis and implementation choices within an already chosen project.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- jjfroehlich/agent-handbook-for-life-science-researchword for word, first seen Oct 9, 2026
Other versions1 project uses a version of this.
Word for word · 1
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Use this skill when the user needs to choose among scientific project ideas, decide whether a research project or hypothesis should …
- Distinctive: line 57
Do not treat exciting, novel, or difficult ideas as automatically worth pursuing.
- Distinctive: line 58
Do not let sunk costs, identity, or external encouragement replace current evidence.
- Distinctive: line 60
Do not treat a failed control, unavailable resource or non-significant comparison as proof that the biological idea is false.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Research Strategy And Project Design
Purpose
Help the user turn unclear research possibilities into sharper questions, tractable project designs, explicit tradeoffs, and near-term decisions.
Use this skill when
Use this skill when the user needs to choose among scientific project ideas, decide whether a research project or hypothesis should continue, design the next decisive scientific test, generate hypotheses, or rebalance a research portfolio.
Do not use this skill when
Keep fixed-project analysis and implementation outside this skill: dataset/features, models, statistical contrasts, genomic windows, diagnostics, plots, protocols and code. A plan or go/no-go request belongs here only when it changes the scientific question, experimental program or project continuation. Earlier strategy work does not keep the skill active for later technical tasks.
Core workflow
- Identify the scientific choice the user is making; use their existing evidence and constraints before asking for missing context.
- Compare the question's knowledge gain and personal interest with practical feasibility and the decision horizon. Give reasons; do not invent numerical scores.
- Distinguish idea generation from testing without imposing a fixed order. Use background, observations or discussion to develop a question; allow results to reshape it and identify discriminating predictions when ready.
- Route to the narrowest playbook needed.
- Name the uncertainty that changes the next commitment. Check a necessary premise early when a credible inexpensive check exists; otherwise compare the real commitment and alternatives.
- Recommend an action proportionate to the evidence. Distinguish failed measurements and imprecise results from evidence against a biological prediction. A pause can leave the question unresolved.
Output formats
Select only the format that fits the requested deliverable; these are patterns, not mandatory response sections.
- Short strategy memo with recommendation, rationale, decisive uncertainty, and next test.
- Project comparison table with the dimensions that change the choice, such as feasibility, knowledge gain, timeline, support and next informative check.
- Hypothesis-generation plan that separates exploratory moves from confirmatory tests.
- Risk register with continuation criteria, failure modes, and review date.
Reference routing
- Open
references/choosing-problems.mdwhen the user is choosing a problem, comparing novelty and tractability, or asking what makes a scientific question worth pursuing. - Open
references/project-triage.mdwhen the user has one or more project ideas and needs a pursue, pause, pivot, or quit decision. - Open
references/risk-and-kill-criteria.mdwhen the user needs a cheap decisive test, continuation threshold, feared control, or bias-resistant review. - Open
references/experiment-types.mdwhen the strategic question depends on what kind of evidence, scale, or experiment type should come next. - Open
references/hypothesis-generation.mdwhen the user is stuck, needs new hypotheses, is interpreting unexpected observations, or wants more creative routes into a problem. - Open
references/handbook-access.mdwhen a fuller explanation or worked example would help with the current task.
Quick checklist
Apply only the checks relevant to the requested decision. Inspect available context first; ask only for missing information about the question, constraints, tools/data, time horizon, career or portfolio stage, or decision that would materially change the recommendation.
Common pitfalls
- Do not treat exciting, novel, or difficult ideas as automatically worth pursuing.
- Do not let sunk costs, identity, or external encouragement replace current evidence.
- Preserve exploratory origins and account for selection before treating a finding as confirmation; a later test cannot retroactively make the discovery prespecified.
- Do not treat a failed control, unavailable resource or non-significant comparison as proof that the biological idea is false.
Quality bar
- Make tradeoffs visible instead of treating every idea as equally good.
- State the conditions that change the recommendation; use quantitative continuation thresholds only when the prediction, assay and user's constraints justify them.
- Separate exciting from executable.
- Return the requested answer; do not require a risk register, worksheet or full review for a small decision.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

