codex
claudecodexDelegate work to the OpenAI Codex CLI — reviews, background tasks, and rescue sessions
13 skills · hooks
Data Request — triage, analyst intake, scope, lookup, map, draft, validate, fix and amend request SQL and Python, and draft analyst-approved release summaries, with shared RDL guardrails and human-confirmed review handoffs
Plugin data-request from marketplace rdl-agent-extensions in nq-rdl/agent-extensions. Commands read from the repo's manifest — 27 days ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.
amendClassify analyst amendments to an operator-run or released RDL extract. Renames, reorders, dropped already-surfaced columns and presentation changes go through query-builder's projection extension points; anything that changes rows or meaning becomes a paste-ready engineer hand-off. Use when an extr
analyseReview a finished SQL file for handoff: read it against its scope, record purpose, inputs, outputs, logic steps, assumptions and limitations in .sqlreview/reviews/<slug>/review.json, with every assumption and limitation confirmed by the human, and render the standardised review.md. Re-running on rev
bootstrapScope a piece of SQL work with the Data Engineer before the SQL is written: an interview that produces .sqlreview/reviews/<slug>/scope.json and the rendered scope.md (intent, inputs, outputs, assumptions, open questions), preserving recorded upstream confirmations. Re-running on an existing scope wa
draftDraft or revise RDL cohort SQL from a defined request and verified source mappings, working autonomously on settled requirements or co-developing unresolved decisions. RDL cohort SQL; verify target engine/version, dataops schema and query-builder column-spec metadata at use time. record_assumption/r
explainWalk a Data Analyst through reviewed SQL, step by step, against its review document, JSON and scope: each logic step with its lines, the assumptions and limitations that govern it (by id), outputs and open questions — pausing for the analyst at every step, then accept for release preparation, send b
fixFix a defect or make a decided pre-release logic change in RDL request SQL and Python: query results, transformations, identifier types, export formatting, keys, joins and filters. Use for a defect with a concrete expected result, released or not, and for a logic change confirmed to come before an e
guardrailsApply advisory RDL cohort-SQL guardrails when drafting, changing, mapping, validating or reviewing request pipelines, SQL and query-builder resolvers. Covers spec composition, indexed filters, source timezones, join evidence and researcher modelling choices, with pointers to authoritative sources. R
intakeInterview the Data Analyst on the research decisions of a data request — cohort and age limits, code-set edges, outcomes and windows, identifiers and presentation, governance and the grain — then fill answers.yaml, write answers.intake.json (schema 1) and, on approval, render the data-analysis-scaff
liftClose out request pipelines: inspect the pinned library, classify lift candidates with the human, file reusable work on the owning library, and record delivered hand-SQL limitations. Revisit released lifts for recurring extracts and candidates in builds not yet delivered, without automatic adoption.
lookupHelp a Data Analyst discover source codes for an order field, clinical event or event set, medication, diagnosis or procedure, or pathology task assay. Choose the lookup family and direction, generate a count-only probe for a human to run, then write a private lookup record for mapping. Never runs a
mapPropose evidence-backed RDL cohort source-table and query-builder resolver mappings from a request, scope, legacy answers.yaml or existing hand SQL; resolve grain, keys, timezone and units before SQL drafting. Supports a partial map. repo: https://github.com/nq-rdl/agent-extensions
releaseDraft the researcher-facing Extraction Summary and Extraction Assumptions / Important limitations for a release from the release tag's own pipeline, outputs and validation, checked against the confirmed scope and .sqlreview reviews. A stale review or conflicting source holds its claim; the analyst a
setupInitialise project scope and review records for the Data Request workflow: create the .sqlreview/ directory with config.json (settings + the shared assumption/limitation definitions) and the scope/review report templates, from the bundled default or from answers to a few questions. Safe to re-run —
triageTriage the RDL service-desk data-request queue: resolve enquiry and issue numbers to approval-ID repositories, check scope, scaffold, branch and library-gap state, classify blockers, and return paste-ready comments with an ordered queue (read-only). Co-development mode works one selected request wit
validateValidate RDL request or cohort SQL against its requirements, column specs and shared guardrails; report technical findings and evidence without creating a confirmed review. repo: https://github.com/nq-rdl/agent-extensions
"${CLAUDE_PLUGIN_ROOT}/hooks/data-request-preflight.sh"
"${CLAUDE_PLUGIN_ROOT}/hooks/data-request-guard.sh"
Delegate work to the OpenAI Codex CLI — reviews, background tasks, and rescue sessions
13 skills · hooks
GitHub workflow — git hooks, changelogs, conventional commits, pull requests, releases, Go CI/CD, and secure GitHub Actions (SHA pinning, OIDC, GitOps delivery)
12 skills
Claude Code — agent-team coordination, hook authoring, and skill-quality auditing
9 skills · hooks
OpenCode development toolkit — author plugins, agents, the SDK, custom tools, skills, governance policies, and delegation harnesses
7 skills · hooks
R — language expertise, package development, CLI apps, CRAN, lifecycle, async, and testing
8 skills
RDL team workflows — repository context, spec-kit delivery, and onboarding
6 skills
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.
1 MCP server
Use Alibaba Open Code Review from local Codex.
2 skills
Claude Code skill for creating and presenting Slidev slidedecks
1 skill
Code intelligence powered by a knowledge graph. Provides execution flow tracing, blast radius analysis, and augmented search across your codebase.
12 skills · 1 MCP server · hooks
daisyUI component library skill
7 skills
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.