azure-id
Community@azure-id
azure-id publishes ORC, a 40-skill orchestration system for Claude Code that drives requirements analysis, parallel task execution, verification, and stacked-PR shipping.
Agent Skills by azure-id
Showing 40 vetted skills indexed across 1 GitHub repositories.
azure-idorc-doc
Write long structured documents like PRDs and TSDs as portable Markdown via section-based orchestration.
azure-idcontext-combiner
Merges multiple confirmed requirement analyses into one deduplicated, conflict-resolved specification.
azure-idorc-pact
Tracks project invariants in a ledger and recomputes their status against code changes.
azure-idorc-brainstorm
Generates and stress-tests idea candidates through structured diverge-converge brainstorming phases.
azure-idorc-pattern
Codifies a project's real code conventions into cached per-language pattern files for ORC executors.
azure-idorc-poly
Plans one change spanning multiple repositories with a frozen interface contract.
azure-idorc-retro
Aggregates ORC behavior traces into calibration reports filed as upstream GitHub PRs.
azure-idorc-judge
Evaluates analysis, plan, and implementation artifacts at three ultra-lane judgment gates.
azure-idorc-export
Export ORC project context into portable AGENTS.md files and import foreign agent context.
azure-idorc-analyze
Analyzes requirements against code with quote-anchored evidence before planning begins.
azure-idorc-aftermath
Grades past shipped runs against git history, traces, and the pact ledger.
azure-idorc-pr-driver
Executes stacked pull request plans with per-layer build, test, and lint gates.
azure-idorc-budget
Forecasts token, dollar, and quota costs of an ORC execution plan from usage transcripts.
azure-idorc-analyze-mini
Analyzes requirements into scope-bounded, code-grounded reports and derived specs in a single pass.
azure-idorc-diy
Gates and dispatches user-compiled custom orchestrator flows built via the orc diy CLI.
azure-idorc-quick
Dispatches user-approved subagents to execute small code changes, bug hunts, and PR comment fixes.
azure-idorc-fast
Implements small coding tasks via a single executor using cached wiki and pattern knowledge.
azure-idorc-handoff
Maps and safely edits non-engineer-owned files with graded checks and undo commands.
azure-idorc-pr-setup
Plans stacked pull request layers and writes a stack-plan.md file without touching git.
azure-idorc-challenge
Grades finished documents and code against a stated goal, then stops for external fixes.
azure-idorc-wiki
Scans a codebase and builds a persistent, evidence-anchored project knowledge base in wiki/ docs.
azure-idorc-verify
Verifies git-modified changes with builds, tests, and severity-classified findings.
azure-idorc-explain
Re-explains the last ORC output in plain language with background context and term definitions.
azure-idorc
Orchestrates multi-task builds through scored parallel subagents from intake to ship.
Frequently Asked Questions About azure-id
FAQPage SchemaWhat tasks can I accomplish with azure-id's ORC skills?▼
ORC covers the full build lifecycle: analyzing documents into code-grounded requirement specs (orc-analyze), planning task graphs (orc-planner), parallel scored execution (orc), review/verify/security passes (orc-review-verify), end-to-end testing (orc-test), and shipping via commits, PRs, or stacked PRs (orc-pr, orc-pr-driver).
Who is the ORC skill system designed for?▼
Software engineers running Claude Code who want subagent-driven builds, plus adjacent roles: orc-handoff lets non-engineers safely edit content files, orc-doc serves PMs writing PRDs and TSDs, and orc-learn onboards developers to unfamiliar features with generated docs.
How do I run ORC lanes in practice?▼
Invoke slash commands inside Claude Code: /orc for the full pipeline, /orc-mini or /orc-fast for lighter lanes, /orc-analyze for requirement analysis, /orc-verify for standalone checks. The orchestrator dispatches pinned subagents (e.g., Opus 5, Sonnet 5) and checkpoints state for fresh-session resume.
What prerequisites do the faster ORC lanes require?▼
orc-fast requires a fresh project wiki (built by /orc-wiki) and a cached language code-pattern (from /orc-pattern); if either is missing it falls back to orc-mini. orc-retro requires the gh CLI or a GitHub MCP for report delivery, and orc-diy requires a CLI-compiled flow config.
Does ORC modify my repository without consent?▼
No. Expensive scans (orc-wiki, orc-pattern) are consent-gated, read-only lanes like orc-verify and orc-aftermath never edit code, orc-pr-setup never touches git history, and orc-export never applies an import without explicit approval. Executors commit only within the orchestrated ship phase.