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 Ahmad-Jaradat-Space/co-scientist-plugin --skill evolving-hypothesesgit clone --depth 1 https://github.com/Ahmad-Jaradat-Space/co-scientist-pluginWrote 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/ahmad-jaradat-space/co-scientist-plugin/evolving-hypotheses)<a href="https://agentmods.dev/skills/ahmad-jaradat-space/co-scientist-plugin/evolving-hypotheses"><img src="https://agentmods.dev/badge/skills/ahmad-jaradat-space/co-scientist-plugin/evolving-hypotheses/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/ahmad-jaradat-space/co-scientist-plugin/evolving-hypotheses"><img src="https://agentmods.dev/badge/skills/ahmad-jaradat-space/co-scientist-plugin/evolving-hypotheses.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 | $0.00099 | $0.01059 |
| Opus 5 | $0.00049 | $0.00530 |
| Sonnet 5 | $0.00020 | $0.00212 |
| Haiku 4.5 | $0.00010 | $0.00106 |
Grade A, and why
evolving-hypotheses 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 9d 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 — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Evolving hypotheses
You are acting as the Evolution agent. Evolution never edits a hypothesis in place. It produces a new hypothesis that records its parents, so both stay in the tournament and the ranking decides which survives.
A new hypothesis that is not recorded does not exist. Call
co-scientist:hypothesis_record with parent_ids set.
Pick the operator
Choose by what is wrong, not by what sounds interesting. Each operator has a reference file with its procedure and its failure mode; read the one you pick.
| Operator | Use when | Reference |
|---|---|---|
combine |
Two hypotheses each explain part of the phenomenon | reference/combine.md |
simplify |
The mechanism has a broken or unnecessary step | reference/simplify.md |
feasibility |
The mechanism is plausible but untestable as stated | reference/feasibility.md |
out_of_box |
The whole framing is stuck and the field's assumptions need dropping | reference/out-of-box.md |
If a review found a fatal flaw in one claim but the rest holds, simplify is
almost always the right operator: cut the broken claim and see whether what is
left still explains the phenomenon.
Workflow
- [ ] 1. Read the parents and their reviews
- [ ] 2. Name what is wrong, in one sentence
- [ ] 3. Pick the operator and read its reference file
- [ ] 4. Write the new hypothesis
- [ ] 5. Check it is actually different
- [ ] 6. Record it with its parents
- [ ] 7. Review it
1. Read the parents
co-scientist:hypothesis_get for each parent. Read the reviews. Evolution
without the reviews just reshuffles words.
2. Name the defect
Write one sentence: what specifically fails. "The receptor direction is reversed." "The experiment does not discriminate." "It needs a cell type the tissue does not contain."
If you cannot write that sentence, do not evolve the hypothesis. Review it first.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago First seen · 105 lines · 99 tokens per session scan A ab77a2e9976d
evolving-hypotheses is a skill published in the GitHub repository Ahmad-Jaradat-Space/co-scientist-plugin (0 stars, last pushed 20d ago), licensed Apache-2.0. It adds 99 tokens to every session and 1,059 once invoked, about $0.0005 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
molecular-dynamics
Run and analyze molecular dynamics simulations with OpenMM and MDAnalysis. Set up protein and protein-ligand systems with PDBFixer, choose force fields and water models (AMBER14, CHARMM36m, ff19SB, GAFF2, TIP3P), solvate and add ions, run energy minimization, NVT/NPT equilibration and production MD on GPU, then…
molfeat
Molecular featurization hub with one consistent interface over 100+ featurizers. Fingerprints (ECFP/Morgan, MACCS, atom pair, topological torsion, Avalon, RDKit, ERG), RDKit and Mordred descriptor sets, pharmacophore and 3D shape descriptors, scaffold keys, and pretrained embeddings (ChemBERTa, ChemGPT, MolT5, GIN…
depmap
Query the Cancer Dependency Map (DepMap) for cancer cell line gene dependency scores (CRISPR Chronos), RNAi DEMETER2 scores, PRISM compound sensitivity, and gene effect profiles across the cell-line panel. Use for identifying cancer-selective vulnerabilities, separating pan-essential genes from selective ones, finding…
glycoengineering
Analyze and engineer protein glycosylation. Scan sequences for canonical N-glycosylation sequons (N-X-S/T with X not proline, including overlapping sites), predict O-GalNAc hotspots, read glycan notation, and reach the curated external tooling (NetNGlyc, NetOGlyc, GlycoShield, GlycoWorkbench, GlyTouCan, GlyConnect).…
pytdc
Use Therapeutics Data Commons through the PyTDC Python package for registry discovery, approved dataset access, task-aware splits (scaffold, cold-start, temporal, combination), evaluator metrics, benchmark groups, and bounded molecular-oracle workflows. Use this skill to find which TDC datasets exist for a therapeutic…
rowan
Rowan is a cloud-native molecular modeling and medicinal-chemistry workflow platform with a Python API. Use for pKa and macropKa prediction, conformer and tautomer ensembles, docking and analogue docking, protein-ligand cofolding, MSA generation, molecular dynamics, permeability, descriptor workflows, and related…