chemistry-research
Use when the work involves molecular structures, reactions, spectra, computational chemistry, crystallography or chemical databases.
Skill by AhmiDarrow
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/ahmidarrow-chemistry-research
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: research
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/chemistry-research
What's inside
- Skill name
chemistry-research- Description
- Use when the work involves molecular structures, reactions, spectra, computational chemistry, crystallography or chemical databases. Covers SMILES/InChI round-trips, RDKit standardisation and fingerprints, NMR/IR/MS interpretation workflows, DFT method and basis-set choice, CIF and R-factors, PubChem/ChEMBL/CSD, and the characterisation standard a journal expects. Hazard, waste and institutional review are part of the procedure.
- License
- Apache-2.0
- Instructions
- 91 lines in SKILL.md · view on GitHub
Files (9)
- SKILL.md
- INDEX.md · references/
- chemical-databases.md · references/
- cheminformatics-rdkit.md · references/
- computational-chemistry.md · references/
- crystallography-and-cif.md · references/
- lab-safety-and-scope.md · references/
- spectra-workflows.md · references/
- structure-representations.md · references/
SKILL.md (first lines)
# Chemistry research Activate `research-method` first: `skill_activate(skill="research-method")`. It owns question framing, evidence standards, citation honesty and how to say "we do not know". Do not restate it. This pack is what chemistry does differently. ## Scope and safety, stated once You do the ordinary published chemistry: literature search, structure and reaction handling, spectral interpretation, computational chemistry, crystallography, purification and characterisation, statistics on assay data, and the hazard assessment that goes with all of it. You decline synthesis routes, scale-up detail, and procurement or precursor-sourcing guidance for explosives, chemical weapons, drugs of abuse and scheduled or watched precursors — including "hypothetical" and "for a novel" framings, and including the last missing step of an otherwise public route. When a request lands there, say so plainly, point at the institutional route (EHS office, chemical hygiene officer, controlled substances licensing, the relevant national scheduling authority), and continue with the legitimate part of the work. Routine hazard work stays in scope and belongs in the procedure, not a footer: pull the SDS for every reagent before writing a step, record GHS/hazard statements, name incompatibilities (oxidiser/fuel, acid/cyanide, water-reactive, peroxide-formers with an expiry date), state containment (fume hood, glovebox, blast shield), quench and waste stream per reagent, and the PPE. See `references/lab-safety-and-scope.md`. ## Structures before anything else A structure that does not round-trip is a bug you will not notice later. Canonicalise once, keep the canonical form, and record what you did: - **SMILES** is not unique unless canonicalised, and canonical form is toolkit-specific — say which toolkit and version produced it. - **InChI / InChIKey** is the identity key for exact-match lookup. The standard InChI deliberately loses information (tautomers are normalised, some stereo layers can be absent); an InChIKey collision is a match on the skeleton layer, not proof of identity. - Strip salts and solvates to a parent for comparison, but keep the original — the salt form changes the assay result. - Check stereocentres and double-bond geometry survived; unspecified stereo is different from racemic, and both differ from a single enantiomer. - Charge, protonation state and tautomer are pH-dependent choices, not facts. State the pH you assumed. Details and the RDKit standardisation pipeline: `references/structure-representations.md`, `references/cheminformatics-rdkit.md`. ## Running chemistry code RDKit, Open Babel, psi4, ASE and pymatgen are not importable inside Remedy — the sidecar excludes the numeric stack. `analysis_env(probe=True)` to see what the project has, then run every calculation through `analysis_run` in the project's own interpreter so the script, its inputs and its outputs are hashed into the ledger. A descriptor table with no run behind it cannot be checked later. ## Claims that need a chemist Retrosynthesis suggestions, reaction-condition predictions, ADMET and docking scores are hypotheses ranked by a model, not results. Present them as ranked suggestions with the model and version named, and say plainly that a chemist decides. Never report a predicted spectrum, yield or property as a measurement. ## Verification — what "done" means here - Every structure round-trips: SMILES -> mol -> canonical SMILES -> InChIKey, compared to the source, with the toolkit version recorded. - Every compound identity claim rests on the characterisation set the journal requires (typically 1H and 13C NMR with assignments and integrations, HRMS within the stated mass accuracy, plus IR/mp/optical rotation as relevant) and purity evidence (HPLC/GC trace or elemental analysis) — not on a single spectrum. - Yields are isolated yields with the limiting reagent named and the scale stated; a crude or NMR yield says so. - Computational results state functional, basis set, solvation model, dispersion correction, convergence criteria and the software version, and the geometry was confirmed a minimum (no imaginary frequencies) or a transition state (exactly one). - Crystallographic claims carry the CIF, R1/wR2, GooF, completeness and the checkCIF alerts, with A/B alerts explained. - `cite_check(resolve=True)` returns PASS before the manuscript is done; `manuscript_build` reports no undefined citations. Read `references/INDEX.md` and pull what you need with `file_read`.
From AhmiDarrow/RemedyAI, Apache-2.0.
More from AhmiDarrow/RemedyAI
- 0.624score
bevy
AhmiDarrow
Build Rust games with Bevy: ECS mental model, plugins, schedules, states, assets, 2D sprites, fast iteration (cargo check, dynamic_linking, opt-level)…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
clinical-research
AhmiDarrow
Human-subject studies run to the standard the field expects: choosing a design that can answer the question, randomisation and blinding, endpoints fixed…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
bioinformatics
AhmiDarrow
Use for sequence and omics work: FASTQ/BAM/VCF/GTF/BED handling and their coordinate conventions, reference builds and liftover, read QC and trimming…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
comfyui
AhmiDarrow
Local ComfyUI image generation end-to-end: from a blank machine (download, install, start, fetch models, workflows) through queue/poll/download and chat embeds.
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
computational-science
AhmiDarrow
Simulation and numerics you can defend: floating-point and conditioning understood before blaming the physics, solvers chosen for the problem, convergence and…
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
stress-suite
AhmiDarrow
Live stress and break suites for Remedy API/desktop — multi-turn storms, concurrent sessions, multi-provider rotation, adversarial break battery.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
soak-product
AhmiDarrow
Full product soak for Remedy — API health, computer-use, plan mode, browser rail, concurrency, docs check, and sign-off artifacts.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
change-safety
AhmiDarrow
Blast-radius protocol before code changes — name the surface, list coupled neighbors, run paired checks, and smoke UI/chrome/messengers that unit tests miss.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
project-etiquette
AhmiDarrow
Default ship discipline for any serious project — blast radius first, then fix, test, update code/docs, build, commit, wait for CI, publish only if green.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
self-dev-loop
AhmiDarrow
Meta-loop for Remedy developing herself — self-inject, change-safety, unit tests, security gauntlet, product soak, stress, then ship gates.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
self-inject
AhmiDarrow
Run one test-gated self-improvement round on the Remedy codebase: draft a change, snapshot the diff, gate it with tests only, apply (restart sidecar / rebuild…
SkillAFreshness: A★ 5Claude Code +9 - 0.616score
webui-parity
AhmiDarrow
When changing Grove, Studio, Settings, or chat chrome, keep WebUI (browser on :7400) at parity with Tauri.
SkillAFreshness: A★ 5Claude Code +9
Signals
- Stars
- 5
- Forks
- 1
- Open issues
- 0
- Last commit (this folder)
- Aug 23, 2026 (2 months ago)
- Last release
- Oct 19, 2026 (10 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- Repository
- github.com/AhmiDarrow/RemedyAI
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from AhmiDarrow
- 0.624score
comfyui
AhmiDarrow
Local ComfyUI image generation end-to-end: from a blank machine (download, install, start, fetch models, workflows) through queue/poll/download and chat embeds.
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
computational-science
AhmiDarrow
Simulation and numerics you can defend: floating-point and conditioning understood before blaming the physics, solvers chosen for the problem, convergence and…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
bevy
AhmiDarrow
Build Rust games with Bevy: ECS mental model, plugins, schedules, states, assets, 2D sprites, fast iteration (cargo check, dynamic_linking, opt-level)…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
bioinformatics
AhmiDarrow
Use for sequence and omics work: FASTQ/BAM/VCF/GTF/BED handling and their coordinate conventions, reference builds and liftover, read QC and trimming…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
clinical-research
AhmiDarrow
Human-subject studies run to the standard the field expects: choosing a design that can answer the question, randomisation and blinding, endpoints fixed…
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
change-safety
AhmiDarrow
Blast-radius protocol before code changes — name the surface, list coupled neighbors, run paired checks, and smoke UI/chrome/messengers that unit tests miss.
SkillAFreshness: A★ 5Claude Code +9
Cite this page
chemistry-research skill. Agent Mart. https://agentmart.dev/skills/ahmidarrow-chemistry-research (last updated 2026-10-24)