opendatahub-io

Use when you need to evaluate CodeRabbit PR comments and fix or reply

Use this skill to write unit tests that strictly conform to the project's existing testing structure, patterns, and style by learning from similar tests before writing anything new.

Review an Architectural Decision Record (ADR) using a team of six specialist reviewer subagents and produce a consolidated report as both PDF and PPTX slide deck. Use this skill whenever the user asks to review, critique, audit, or get feedback on an ADR,…

Use this skill to analyze context sufficiency for documentation generation. Reads workspace/context-package.json and produces workspace/gap-report.json with severity-rated gaps and a proceed/gather-more/stop recommendation.

Use when you need to gather context for a Jira ticket or PR. Resolves ticket metadata, clones relevant repos, collects candidate files, runs filtering pipeline, and produces workspace/context-package.json.

Use when you need to generate AsciiDoc documentation modules from gathered context. Reads context package and gap report, generates content, then self-validates with iterative correction (up to 3 retries). Produces generated files and…

Use this skill to orchestrate the full documentation pipeline. Sequences doc-gather, doc-gap, doc-validate, doc-review, and doc-generate skills based on the requested pipeline mode.

Use this skill to produce a STRAT-level documentation plan. Traverses a strategic initiative's child epics and stories to identify what documentation is needed, what type, and at what priority.

Write Markdown DITA topics and maps, then convert to DITA XML using DITA-OT with the org.lwdita plugin. Use when asked to create DITA documentation, write Markdown DITA topics, generate concept/task/reference content, create DITA maps, or convert Markdown to…

Check that all environment variables and CLI tools required by the docs-skills plugin are available. Use when setting up a new environment, troubleshooting missing prerequisites, or before running a workflow for the first time.

Technical accuracy review with optional code-learner claim validation. Iteration logic owned by the orchestrator.

Unified interface for GitHub PRs and GitLab MRs: read PR/MR data with file filtering, list changed files, get review comments, post inline review comments, reply to existing comment threads, extract line numbers from diffs, validate comments, auto-detect…

Read and analyze JIRA issues from Red Hat Issue Tracker. Use this skill to fetch issue details, search issues, extract comments and discussions, categorize issues (bugs, CVEs, features, stories), analyze custom fields (release notes, fix versions), retrieve…

Documentation workflow orchestrator. Reads the step list from .agent_workspace/docs-workflow.yaml (or the plugin default). Runs steps sequentially, manages progress state, handles iteration and confirmation gates. Claude is the orchestrator — the YAML is a…

Write documentation from a documentation plan. Dispatches the docs-writer agent. Supports AsciiDoc (default) and MkDocs formats. Default placement is UPDATE-IN-PLACE; use --draft for staging area. Also supports fix mode for applying technical review…

Score documentation quality and intent alignment using LLM judge agents (Opus). Dispatches two judge agents in parallel (doc_quality, intent_alignment), extracts specific gaps from AC coverage analysis, and cross-references against scope-req-audit evidence.…

Guides Konflux CI/CD pipeline onboarding for new components — Jira tracking epic, upstream ODH Dockerfile, downstream RHOAI Dockerfile, standalone deployment manifests, OpenShift CI, dashboard-operator registration, and DevOps coordination. Operator…

Scaffold a new federated module under packages/. Runs mod-arch-installer, allocates ports, registers the feature flag and SupportedArea in the host, and verifies the build. Standalone manifests and operator registration are deferred to /konflux-onboarding…

Debug envtest integration test failures in the Dashboard Module Controller. Analyzes test output, identifies failure patterns, reads relevant test code, and suggests fixes. Use when envtest tests fail locally or in CI.

Builds and deploys the gen-ai standalone container to a personal OpenShift cluster for testing. Handles frontend build, Go BFF cross-compile, Podman image build/push to quay.io, and deployment patching. Supports both initial deploy and fast iteration (rebuild…

Parses Jest coverage reports to find uncovered lines, then generates targeted unit tests to increase coverage for a specific package or file. Use when you want to improve test coverage for a particular area of the codebase.

Convert fine-grain Cypress mock tests to Jest unit tests based on CI execution time

Sync upstream changes from a local repository clone. Pass a package name, local repo path, and branch as arguments. Supports cherry-pick (--commit=SHA) and range (--up-to=SHA) modes.

Sync upstream changes for a package and open a PR. Pass a package name as the first argument (e.g. model-registry or notebooks) or be prompted to choose. Optionally pass a PR URL to do a temporary test sync.

Reviews strategy features for architectural correctness — dependencies, integration patterns, component interactions.

Reviews strategy features for technical feasibility, implementation complexity, and effort estimate credibility.

Reviews Initiatives for technical feasibility, blockers, and dependency realism.

Reviews RFEs for technical feasibility, blockers, and alignment with technical strategy.

Review and improve Initiatives. Accepts one or more Jira keys (e.g., /initiative-review RHOAIENG-12345) to fetch and review existing Initiatives, or reviews local artifacts from /initiative-create.

Review and improve RFEs. Accepts one or more Jira keys (e.g., /rfe.review RHAIRFE-1234 RHAIRFE-5678) to fetch and review existing RFEs, or reviews local artifacts from /rfe.create. Runs rubric scoring, technical feasibility checks, and auto-revises issues it…

Split oversized RFEs into smaller, right-sized RFEs. Accepts one or more IDs (e.g., /rfe.split RHAIRFE-1234 RHAIRFE-5678). Runs non-interactively — decomposes, generates new RFEs, reviews them, self-corrects, and checks coverage.

Validates whether an Initiative aligns with its parent RHAISTRAT Outcome. Checks objective advancement, scope consistency, and success criteria contribution.

Create and protect GitLab branches. Use when the user asks to create a branch, protect a branch, or set up branch protection rules on GitLab. Uses the glab GitLab CLI. The script takes three required positional args in order: repo, branch-name, ref (no --ref…

Troubleshoot and analyze logs from AWS CloudWatch Logs. This skill should be used when the user asks to investigate logs, troubleshoot application issues, query log groups, analyze error patterns, or perform log analysis for machines writing to AWS…

Code review a GitLab merge request using multiple specialized agents with confidence-based scoring to filter false positives. Use when the user asks to review an MR, do a code review, or check changes in a merge request.

Analyze GitLab CI/CD job failures, parse logs, identify error patterns, and troubleshoot pipeline issues. This skill should be used when the user asks to investigate failed jobs, debug pipeline failures, analyze job logs, compare job runs, or generate failure…

Check Google Calendar for team availability and upcoming events. Use when the user asks about free/busy status, team schedules, upcoming events, or calendar availability. Uses the Google Workspace CLI (gws). Requires prior authentication via `gws auth login`…

Scan Google Docs shared with you for action patterns (e.g. "create a card", "action item", "open a ticket") in both content and comments, extract them as structured data, and create Jira issues. Use when the user asks to scan docs for actions, find TODOs in…

Read and search Google Drive files (Docs, Sheets, Slides, and plain files). Use when the user asks to list, search, or read the content of any file in Google Drive, including Google Docs and Google Sheets. Uses the Google Workspace CLI (gws). Requires prior…

Search, read, create, and add diagrams to Google Slides presentations. Use when the user asks to find a presentation, read slide content, create a new Google Slides deck, or add visual diagrams to existing slides. Uses the Google Workspace CLI (gws). Requires…

Generate a test plan from a strategy (RHAISTRAT or RHOAIENG issue), with optional ADR for extra technical depth. Use when starting test planning for a new RHOAI feature with a defined Jira strategy.

Generate one complete test file with all functions for assigned test cases, including quality scoring and auto-revision

Publish test plan artifacts to GitHub — creates a branch, commits all artifacts, and opens a PR with optional reviewer assignment. Use after test plan review to make artifacts available for team collaboration and formal review feedback.

Reviews a generated test plan for completeness, consistency, and quality using a 5-criteria rubric. Scores, auto-revises, and re-scores (max 2 cycles). Use for automated quality assessment and iterative improvement of generated test plans.

Score an existing test plan using the quality rubric without triggering auto-revision. Use for standalone quality assessment of test plans or evaluating test plans created outside the automated generation pipeline.

Score generated test function code for completeness, quality, and convention adherence using a 5-criteria rubric. Use for validating generated test code quality before including in the final implementation.

Generate individual test case files from an existing test plan. Use after test plan approval to generate individual TC specifications with preconditions, steps, and expected results organized by category and priority.

Generate executable test automation code from test case specifications (default target: opendatahub-tests; override with --target-repo). Skips TC-UI-*. Use after test cases are reviewed to create production-ready pytest code that follows repository…

Combine multiple component architecture summaries into a platform-level architecture document. Use after generating component summaries to create a wholistic platform view.

Analyze an ODH/RHOAI component repository and generate an evidence-backed architecture summary.

Use when the AIPCC base image repository has changed and the overlay file overlays/0017-aipcc-base-images.md needs to be refreshed with current accelerator support information, versions, or architecture details.

Use when the rhai/pipeline repository has changed and the overlay file overlays/0020-rhai-pipeline.md needs to be refreshed with current product version, published variants, collection matrix, or Pulp publishing details.

Use when the rhaiis/pipeline repository has changed and the overlay file overlays/0021-rhaiis-pipeline.md needs to be refreshed with current vLLM versions, variant matrix, or package content details.

Use when the AIPCC wheels builder repository has changed and the overlay file overlays/0019-wheels-builder.md needs to be refreshed with current builder version, variant matrix, pipeline-API contract, or plugin system details.

Discover platform components by exploring breadcrumbs (installers, operators, dependencies) in checkouts directory. Outputs component-map.json for platforms without manifest scripts.

Generate architecture diagrams (Mermaid, C4, security network diagrams) from GENERATED_ARCHITECTURE.md. Creates visual representations for different audiences - developers, architects, security teams.

Capture a PyTorch build-deliverable request (build or sub-version build). Asks clarifying questions, then produces a structured artifact with YAML frontmatter. Use when a consumer needs a new build target.

Validate a PyTorch build-deliverable request — a definition-of-ready check. Confirms the request is complete, well-formed, and internally coherent, then writes a review file. Run after /torch.build.create or against an existing AIPCC- key.

Submit reviewed PyTorch build-deliverable requests to Jira. Creates AIPCC tickets, applies provenance labels, renames artifact files to the Jira key. Use after /torch.build.review when status is Ready.

Validate an internal ask — a definition-of-ready check. Confirms the request is complete, well-formed, internally coherent, and carries the governance signal (named approver + approval status) submission requires, then writes a review file.

Capture a team ask — a cross-cutting request, a dispute, or an unblock-model-validation escalation routed to the architect for prioritization. Asks clarifying questions, then produces a structured artifact with YAML frontmatter.

Validate a team ask — a definition-of-ready check. Confirms the request is complete, well-formed, internally coherent, and carries the technical context the architect needs to route it, then writes a review file.

Submit reviewed team asks to Jira. Creates AIPCC tickets, applies provenance labels, renames artifact files to the Jira key. Use after /torch.team.review when status is Ready.

Capture an internal ask — a request from another AIPCC team or a PyTorch working group, including RFC handoffs and backports. Asks clarifying questions, then produces a structured artifact with YAML frontmatter.

Generate eval.yaml for the agent eval harness. Two modes - (1) Skill-based - examines SKILL.md, sub-skills, scripts, test cases to verify implementation quality, OR (2) Prompt-based - tests agent capabilities using custom analysis prompts (documentation…

Generate evaluation test cases for an eval.yaml. Sources cases per generation.strategy - skill analysis (default), synthetic LLM generation from generation prompts (documentation and agent-capability evals), or MLflow production traces. Bootstraps a starter…

Execute an evaluation against test cases (skill or prompt mode), score with judges, and report results. Requires eval.yaml (generated by /eval-analyze). Use when the user wants to test a skill, run eval, benchmark, compare models, detect regressions, check…

MLflow integration for evaluation — sync datasets, log run results, push/pull feedback between the harness and MLflow traces. Use when the user wants to log eval results to MLflow, sync test cases to MLflow datasets, connect judge scores to traces, pull…

Run Design-of-Experiments (DoE) evaluations with ANOVA over a matrix of agent configurations — comparing models, thinking-effort levels, prompts, or other factors across shared test cases, with repeated-measures / mixed-effects statistics that account for…

Evaluate the full harness configuration as a system. Scans all skills, commands, CLAUDE.md, and hooks for redundancy, overlap, type misclassification, and structural issues. Produces an informational report with restructuring suggestions. Use when the user…

Compare evaluation results across multiple models or runs. Takes a directory of eval run artifacts (summary.yaml, run_result.json, HTML reports) and produces a tabbed HTML comparison report with model cards, quality/cost tables, per-case breakdowns, and…

Automated skill improvement loop. Runs eval, identifies judge failures, reads traces and rationale, edits the SKILL.md to fix issues, re-runs to verify, and checks for regressions. Use when the user wants to automatically improve a skill based on eval…

Sign off on a CI-approved strategy — pushes content and adds strat-creator-human-sign-off label. For rubric-pass strategies only.

Refine a single strategy — add the HOW, dependencies, impacted teams/components, and non-functional requirements. Requires a strategy key argument.

Create strategies from approved RFEs by cloning them to RHAISTRAT in Jira, or guiding the user through manual cloning.

Push a locally-refined strategy back to Jira and resubmit to CI. Works for both needs-attention and rubric-pass strategies.

Adversarial review of a single refined strategy. Scores against rubric, then runs independent forked reviewers for detailed prose. Requires a strategy key argument.

Pull a RHAISTRAT issue from Jira into the local/ workspace for human review. Only works on post-CI strategies.

Reviews strategy features for testability — are acceptance criteria measurable, are edge cases covered, can this be validated?

Reviews strategy features for architectural correctness — dependencies, integration patterns, component interactions.

Use when a Python package needs to be onboarded into the fondue monorepo (builder/ and/or rhai-pipeline/). Analyzes package information, AI packaging analysis, and optional build failure details to configure the package, then creates one or two git commits…

Use when you need a concise 2-3 line plain-prose summary of a package onboarding outcome for reporting or ticket updates. Summarizes the build result, root cause of any failure, and key next step.

Use when a Python package build has failed in the RHAI pipeline and you need root-cause analysis. Reads the failure log excerpt and package context, diagnoses the root cause, and writes a structured analysis report.

Use when you need to extract actionable requirements, decisions, and blockers from Jira ticket comments for a Python package onboarding request. Produces a structured summary with four sections.

Use when a Python package needs a security audit before RHAI pipeline onboarding. Triages pre-computed scan outputs and optional source URL, then writes a Markdown report and machine-readable risk verdict.

Use when a Python package needs enterprise packaging investigation before RHAI distribution — covering build system, native dependencies, platform support, and packaging strategy. Produces a Markdown analysis and a machine-readable complexity verdict.

Use when assessing whether a Python package's license permits redistribution in Red Hat's AI package distribution pipeline (RHAI). Reads package context, identifies the SPDX license, and writes a structured compatibility verdict.

Use when probe tests need to be created for a Python package in the wheels-test repository. Analyzes package information, AI packaging analysis, and Jira context to determine test cases, then creates a git commit.

Autonomously resolve Python CVE tickets from Jira - groups CVEs by package+branch, creates PRs, monitors Konflux builds, and updates tickets. Requires Atlassian MCP.

Provision and deprovision ROSA HCP clusters on the shared RHOAI AWS account (rh-aws-saml-login, org 7081269). Covers cluster create, G5g pools (prefer 2xlarge), GPU Operator, namespace pull-secret for quay.io/rhoai, pool resize, bring-up/teardown timing, cost…

Provision ephemeral OpenShift and ROSA test clusters via the Slack cluster-bot app (ci-chat-bot) — launch/rosa create/list/done/auth commands, capability matrix (arm64, GPU, IBM Power/Z, machine flavor), ready-to-use recipes, and known limitations (no…

Validate ARM64 CUDA notebook/runtime images end-to-end — Konflux manifest audit (skopeo arm64), testcontainers CPU smoke (podman --platform linux/arm64), GPU smoke on ROSA G5g/T4G or rdu2 GH200 Grace Hopper (Pod + oc exec), ODH tests/manual GPU notebooks…

Summarize notebook matrix CI runs into one evolving GitHub PR comment using GitHub Actions and pull request MCP data. Use when triaging build-notebooks failures, summarizing matrix progress, or updating a living CI status comment.

Review pull requests in the notebooks repository using GitHub MCP tools and the project review standards. Use when reviewing a PR, when the user asks for automated PR feedback, or when the workflow invokes Antigravity for GitHub review.