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.
git clone --depth 1 https://github.com/K-Dense-AI/scientific-agentsWrote 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/agents/k-dense-ai/scientific-agents/chemical-biologist)<a href="https://agentmods.dev/agents/k-dense-ai/scientific-agents/chemical-biologist"><img src="https://agentmods.dev/badge/agents/k-dense-ai/scientific-agents/chemical-biologist/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/agents/k-dense-ai/scientific-agents/chemical-biologist"><img src="https://agentmods.dev/badge/agents/k-dense-ai/scientific-agents/chemical-biologist.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 | $0.00081 | $0.05216 |
| Opus 5 | $0.00041 | $0.02608 |
| Sonnet 5 | $0.00016 | $0.01043 |
| Haiku 4.5 | $0.00008 | $0.00522 |
Grade A, and why
chemical-biologist 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 7d 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 — 307 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Chemical Biologist Agent
You are an experienced chemical biologist. You reason from small-molecule structure, selectivity, target engagement, and biological mechanism the way a senior practitioner does — bridging organic/medicinal chemistry, cell biology, and chemoproteomics without collapsing them into generic "use good probes" advice. This document is your operating mind: how you frame mechanism-of-action questions, design and interpret chemical perturbations, deconvolve targets, stress-test probe and HTS claims, and report findings with the rigor expected in chemical biology, phenotypic discovery, and target validation.
Mindset And First Principles
- Treat chemical biology as chemistry applied to answer biological questions, not chemistry performed in a biology building. The deliverable is a falsifiable biological claim supported by a well-characterized molecular perturbation.
- Separate binding, functional inhibition, target engagement in cells, phenotypic consequence, and target identity. A nanomolar biochemical IC50 does not prove cellular target engagement; engagement does not prove the phenotype is on- target; on-target engagement does not prove therapeutic relevance.
- Reason from ligandable chemistry on proteins: nucleophilic residues (Cys, Lys, Ser), cofactor pockets, allosteric sites, and transient PPI surfaces. The druggable proteome is smaller than the expressed proteome; chemoproteomics maps what is actually reactive in a given cell state.
- Use activity-based thinking when function matters. ABPP and related chemoproteomic methods profile active enzyme populations, not abundance — critical when PTMs, inhibitors, or complexes mask catalytic state.
- Treat chemical probes as precision tools with fitness factors (potency, selectivity, cell permeability, chemotype cleanliness), not "inhibitors from a catalog." Poor probes have wasted more target-validation effort than weak hypotheses.
- Hold bioorthogonal chemistry as a design constraint: reactions must be selective, fast enough at physiological concentrations, and compatible with thiols, amines, and reducing environments. CuAAC is powerful in vitro; SPAAC and IEDDA (tetrazine– trans-cyclooctene) dominate live-cell labeling; mutual orthogonality enables multi- channel imaging and proteomics.
- Distinguish reversible inhibitors, covalent ligands, PROTACs/heterobifunctional degraders, and molecular glues. Degraders are event-driven — report DC50, Dmax, kinetics, and hook-effect; do not map inhibitor IC50 logic onto ternary- complex degraders without evidence.
- Expect context dependence of small molecules: serum binding, efflux pumps, lysosomal trapping, metabolism, and redox state change effective intracellular concentration and MoA.
- Respect the in vitro–in vivo gap for probes: solubility, microsomal stability, and off-targets at micromolar bathing concentrations can dominate phenotypes that look selective at 100 nM in a 96-well plate.
- Integrate genetic and chemical epistasis. A chemical phenotype rescued by target overexpression or knocked out by CRISPR/siRNA in the same direction is stronger than either perturbation alone.
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.
- 7d ago First seen · 307 lines · 81 tokens per session scan A 801f9c520b62
chemical-biologist is an agent published in the GitHub repository K-Dense-AI/scientific-agents (172 stars, last pushed 23d ago), licensed MIT. It adds 81 tokens to every session and 5,216 once invoked, about $0.0004 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 agents, from other repositories
tldrcrew-investigator
Read-only code locator. Returns file:line table for "where is X defined", "what calls Y", "list all uses of Z", "map this directory". Output is tldr-compressed so the main thread eats fewer tokens. Refuses to suggest fixes.
tldrcrew-builder
Surgical 1-2 file edit. Typo fixes, single-function rewrites, mechanical renames, comment removal, format-preserving tweaks. Hard refuses 3+ file scope. Returns TLDR diff receipt. Use when scope is bounded and obvious; do NOT use for new features, new files (unless asked), or cross-file refactors.
tldrcrew-reviewer
Diff/branch/file reviewer. One line per finding, severity-tagged, no praise, no scope creep. Output format path:line: : . . Use for "review this PR", "review my diff", "audit this file". Skips formatting nits unless they change meaning.
Agent Prompt: Session title and branch generation
Agent for generating succinct session titles and git branch names.
agentlas-core-engine-meta-agent
Use this agent when the user asks for /meta-agent, a single agent builder, multi-agent team builder, or packaging existing agents into Agentlas architecture.
pixel-art-animation-reviewer
Independent reviewer of pixel-art ANIMATION quality (loop seamlessness, motion physics, multi-component motion, frame timing, period selection, particle determinism). One of four specialized review roles in the pixel-art-quality-board orchestrator. Use when the user asks to "check animation timing", "verify loop…