Claude Skills by monarch-initiative

Query the AOP-Wiki XML export — Adverse Outcome Pathways, Key Events, and Key Event Relationships — through the installable aop-wiki-cli module. Use when asked to find the AOP/KE/KER for a mechanism, stressor, or adverse outcome; to run or write an AOP-Wiki term search config; to pull KER weight-of-evidence, empirical support, or evidence tables; to get event completion/integration rankings; or to refresh the AOP-Wiki XML snapshot.

Use ONLY when the user explicitly asks to orchestrate parallel external agents (Codex or Claude Code) in tmux sessions via tp (tmux-pilot). Or when a user tells you that you're the boss or orchestrator. You can also use this if you are an operclaw agent. NEVER auto-invoke for a generic "do this in parallel" request. For in-process subagents, use superpowers:dispatching-parallel-agents instead.

Skill for curating cancer and neoplastic disease entries in the dismech knowledge base. Use this skill when creating or enhancing cancer entries, adding oncogene/tumor suppressor pathophysiology, curating histopathology findings, adding targeted therapy information, and linking genetic drivers to treatment responses. Covers fusion genes, two-hit hypothesis, oncogene addiction, and precision oncology concepts.

Use when claiming the next disease to curate in dismech. Two-phase pick — open `claim`-labelled issues for what is already taken, then the `stubs/` queue for what is left — then files a `Curate <label> (MONDO:NNNNNNN)` claim issue assigned to the current GitHub user. Accepts an optional integer 1–8 to claim N diseases at once. The skill should also start the curation process.

Skill for collecting recent clinical Practice Guideline citations from PubMed for dismech disorder entries. Use this skill when building or refreshing a care-guideline citation set — searching PubMed by disorder for Publication Type "Practice Guideline" within a recent time window, prioritizing disorders by guideline availability, and exporting a tab-delimited citation table for downstream phenotype/content gap mining (as was done for Fanconi anemia).

Generate dismech definitions from OHDSI/ATLAS cohort definitions or other computable phenotype logic. Use when converting OMOP cohort JSON, drafting PheKB-/OHDSI-style phenotype algorithms, or mapping FHIR/CQL/OMOP rules into dismech `definitions` blocks.

Skill for creating a new mechanism module in kb/modules/ (a conserved pathological process that recurs across disorders). Use when the user asks to create/curate a module, add an Xogenesis (pathological-structure-formation) module, or factor a recurrent mechanism out of several disorders. Covers the module schema shape, the trigger→consequence node chain, the treatment target_mechanisms drug pattern, the Xogenesis open-ontology anchor convention, evidence discipline, validation, and registrat...

Create, edit, review, or audit dismech disease groupings in kb/groupings/*.yaml. Use for Grouping records, member unions, grouping_basis and grouping_rationale, membership_criteria boolean logic, criteria_semantics, differentiating_mechanisms, grouping foreign keys, ontology-closure audits, or grouping validation and rendering. Do not use for a Disease entry's classifications block; use disease-classification for that.

Use when asked to curate the next N disease curation issues already assigned to the current GitHub user. Filters open issues to those where the user is the sole assignee and no PR exists yet, picks the first N, then dispatches /curate in parallel for each. Accepts a positional integer N between 1 and 8.

Skill for populating the `classifications` top-level block of a dismech Disease entry. Covers Harrison's Part assignment, mechanistic nosology, lysosomal storage, IUIS immunodeficiency, channelopathy, and ICD-O morphology fields, with a lookup table from common clinical phrasing to controlled-vocabulary keys.

Mine Disease Trajectories (DT/DisTraj) outputs for comorbidity/trajectory candidates, including parsing DT JSON/TSV, extracting directed pairs, filtering by sex or significance, and mapping signals into dismech comorbidity YAML.

Skill for analyzing and improving compliance in the dismech knowledge base. Use this skill when checking disorder file completeness, identifying missing fields (ontology terms, evidence, descriptions), understanding weighted priority scoring, and systematically improving knowledge base coverage.

Instructions for reviewing a dismech PR, in particular PRs relating to disorder curation, either creating new dismech entries or updating existing ones.

Add, validate, repair, or review evidence references and exact-quote snippets in dismech KB YAML and deep-research reports. Use for PMID, DOI, NCT, ICTRP, or structured-source evidence; reference-cache generation; snippet failures; title snippets; bracket normalization; deep-research citation validation; Named Entity Confusion preflight; evidence_source classification; and final evidence checks before a PR.

Select, add, validate, review, or repair ontology bindings and derived term caches in dismech. Use when changing term IDs, canonical labels, preferred_term values, enum meaning mappings, phenotype, cell-type, biological-process, disease, anatomy, exposure, or treatment annotations; choosing an OAK adapter; resolving label or dynamic-enum failures; handling ECTO/XCO terms; or diagnosing cache integrity and ordering problems.

Write the prose of a GitHub PR body, issue comment, review body, or review reply in this repo. Use before posting anything to GitHub — opening a PR, summarizing or triaging an issue, replying to a reviewer, reporting what a scan found, or explaining why you did not make a change. Covers audience calibration, leading with the finding, and the abstraction and hedging patterns this repo has accumulated. Not for YAML `description` / `explanation` / `notes` prose, which is held to a different stan...

Skill for initiating new disorder YAML files in the dismech knowledge base. Use this skill when the user asks to create a new disorder entry. Also useful for enhancing existing entries.

Assess how much evidence AOP-Wiki actually carries for a Key Event Relationship before curating from it. Use when judging whether a KER, or a set of them, is backed by literature; when reading `completion_score`, `weight_of_evidence`, `empirical_support`, or `oecd_status` off a KER or its parent AOP; when deciding how much deep-research effort an AOP-derived causal edge will cost; or when auditing an existing dismech edge whose provenance traces to AOP-Wiki.

Skill for curating microbiome-related pathophysiology in the dismech knowledge base. Use this skill when adding dysbiosis mechanisms, ecological concepts (Anna Karenina, keystone taxa, colonization resistance), SCFA/metabolite pathways, and linking microbial ecology to disease pathophysiology. Covers IBD, C. diff, obesity, and other microbiome-associated conditions.

Turn a set of AOP-Wiki Key Events into a verified set of dismech causal edges. Use when an issue, pilot mapping, or project page names Key Events (especially Molecular Initiating Events) with no dismech counterpart and you need to find what each one leads to, judge whether the AOP-Wiki record supports the claim, and take the surviving edges through evidence verification. Covers the KER lookup, relevance filtering, chain assembly, and the evidence triage that decides how much literature work e...

Project management skill for markdown-based projects with GitHub Project sync. Use when: (1) viewing/managing projects in projects/ folder, (2) syncing project checkboxes to GitHub Projects, (3) picking up next work item from a project, (4) updating project status after completing work, (5) generating project dashboards. Replaces the older /pm command with enhanced GitHub integration.

Review one provider-generated mechanistic-hypothesis exploration report, including its dataset use, analyses, and artifacts, or reconcile multiple separately assessed reports for one hypothesis. Use when assessing or comparing Biomni, OpenScientist, Kosmos, or another provider's hypothesis report or computational research run.