By HigorAlves
Personal Claude Code plugin orchestrating the plan → debug → verify → ship loop with curated skills and slash commands.
Answer reviewer comments on YOUR open PR — categorize unresolved threads, push fixes, and post replies. Workspace-aware across linked PRs.
Author an Architecture Decision Record (ADR) — produce a numbered docs/adr/NNNN-*.md following the orc:adr-writing template. Also handles supersession and status transitions on existing ADRs.
Clean up after completed or abandoned orc sessions — removes workspace state, registry entry, git worktree, and optionally the merged branch. Destructive; always previews and asks first. Workspace-aware.
Review someone else's open GitHub PR — posts a real PR review with line-anchored inline comments and suggestions. Verdict computed mechanically from severities; mandatory preview gate before posting. Workspace-aware across linked PRs. For your OWN working diff, prefer the bundled /code-review instead.
Systematic root-cause investigation, then fix with TDD — never proposes a fix before finding the cause. Writes a resumable diagnosis to .orc/ state plus a regression test; --jira <KEY> links a ticket. Workspace-aware across sibling repos. For a quick one-off look without session state, prefer the bundled /debug.
Applies a defined list of code changes, runs the project's test suite, and reports a diff. Used by /orc:address to execute reviewer-requested fixes, or any context where a list of changes is already decided and just needs to be made cleanly.
Long-running root-cause investigation for hard bugs and unexpected behavior. Use when a bug needs disciplined isolation — reproduction, hypothesis, instrumentation, regression-testing — before any fix is attempted. Maintains an isolated context and produces a written diagnosis the implementing engineer can act on.
Provisions a fast, reproducible dev environment for QA or standalone use — detects existing Docker setup (compose > devcontainer > Dockerfile), generates a minimal compose when none exists (hybrid by default: services in containers, app on host), boots with healthcheck-gated readiness, and writes docker-env-state.json + an evidence packet to .orc/<branch>/files/env/. Falls back to host-mode boot when Docker is unavailable — never hard-blocks QA. Dispatched by /orc:env, /orc:qa Phase 4, and /orc:flow Phase 6.
Senior-developer agent that implements a defined slice list from a plan + failing test(s). Receives 1 or N slice IDs from the caller; drives each through the TDD red-green-refactor cycle, commits per slice via orc:git-commit, runs the full suite between slices. Default executor in /orc:flow Phase 5 (single instance for sequential slices, multiple parallel instances for parallel-safe slices). Also dispatched by /orc:fan-out for plan-slice-shaped tasks. Escalates back to the user when a slice is ambiguous, requires a new dependency, or can't be made green after a bounded number of attempts.
Reviews someone else's open GitHub PR end-to-end. Fetches diff via gh CLI, walks every changed file, and returns structured findings as JSON for the /orc:code-review command's posting layer to convert into inline GitHub PR comments. Used by /orc:code-review.
Author and update Architecture Decision Records (ADRs) — the durable record of a decision and its trade-offs, in docs/adr/NNNN-*.md. Use to lock in an architectural decision or to supersede an ADR.
Browser automation CLI for AI agents; use for web tasks - navigate, click, fill forms, screenshot, scrape, test web/Electron apps.
Implement Clean, Hexagonal (Ports & Adapters), and Domain-Driven architecture patterns. Use when designing service layers, refactoring monoliths into bounded contexts, or fixing layer dependency cycles.
Ultra-compressed, signal-only pull request descriptions. Use when opening a PR or when the user says "write a PR description", "PR body", "draft the PR", or invokes /caveman-pr.
Ultra-compressed code review comments — one line per finding: location, problem, fix. Use when the user says "review this PR", "code review", "review the diff", "/review", or invokes /caveman-review.
Executes bash commands
Hook triggers when Bash tool is used
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Executables (bin/) — files in this plugin's bin/directory are added to the Bash tool's PATH while the plugin is enabled.
This plugin requires configuration values that are prompted when the plugin is enabled. Sensitive values are stored in your system keychain.
pr_size_budgetSoft LOC budget per PR (additions+deletions, post-exclusion) enforced by the ship/flow/stack-pr gate
${user_config.pr_size_budget}skip_tool_checkSkip the SessionStart CLI dependency pre-flight
${user_config.skip_tool_check}protected_branchesComma-separated branch names guarded by the confirm-to-commit hook
${user_config.protected_branches}No model invocation
Executes directly as bash, bypassing the AI model
No model invocation
Executes directly as bash, bypassing the AI model
Runs pre-commands
Contains inline bash commands via ! syntax
Runs pre-commands
Contains inline bash commands via ! syntax
"Zug zug."
Let the orcs do the work.
A personal Claude Code plugin for the senior-developer day — plan → debug → verify → ship — with curated skills, composite commands, and a senior-dev agent that writes the code while you mind the gates.
orc is a personal-workflow plugin: 57 curated skills, 21 composite slash commands, 12 specialist subagents, and 6 hook scripts that quietly enforce discipline (no commits to main, no AI-attribution trailers in commits/PRs, dependency pre-flight check, core rules injected at every session start). One umbrella command — /orc:flow — drives the full feature lifecycle from "I want to do X" to "PR merged" with orc-implementer writing the code slice-by-slice in between.
It exists for one reason: every time a senior developer sits down to work, they should already know how the next hour goes — write the plan, watch the test fail, fix the cause (not the symptom), verify with evidence, ship the PR. orc encodes that loop.
orc maps the senior IC / tech-lead / architect day to a small set of composite commands. Most work fits this loop:
flowchart LR
plan["/orc:plan"] --> start["/orc:start"] --> impl["implement"] --> qa["/orc:qa"] --> ship["/orc:ship"] --> cleanup["/orc:cleanup"]
cleanup -.->|"interrupted? /orc:resume"| plan
cleanup -.->|"need status? /orc:status"| plan
Outside the loop — reach for these directly when the situation isn't a fresh feature pipeline:
| Situation | Command |
|---|---|
| Debugging a known bug | /orc:debug |
| Needing the app running (QA or manual poking) | /orc:env |
| Reviewing someone else's PR | /orc:code-review |
| Responding to your PR's review comments | /orc:address |
| Authoring a Product Requirements Document | /orc:prd |
| Authoring a Technical Requirements Document | /orc:trd |
| Locking in an architectural decision | /orc:adr |
| Proposing a system design before code | /orc:rfc |
| Writing an incident postmortem | /orc:postmortem |
| Bootstrapping a new package/service | /orc:scaffold |
| Parallel-dispatching N independent tasks | /orc:fan-out |
| Filing/linking a Jira ticket from the terminal | /orc:jira |
Or skip the per-phase invocations and use /orc:flow to drive the whole loop — gates at every phase, autonomous implementation in between via orc-implementer.
| You have ... | Read |
|---|---|
| A whole feature/bug to drive end-to-end | examples/00 — End-to-end with /orc:flow |
| A reproducible bug or failing test | examples/01 — Fixing a bug |
| Just had a production incident | examples/02 — Writing a postmortem |
| A new feature to ship | examples/03 — Adding a new feature |
| A new package/service or doc gap | examples/04 — Writing documentation |
| A PRD from PM | examples/05 — Handling a PRD |
| Someone else's open PR | examples/06 — Reviewing someone's PR |
| Reviewer comments on your PR | examples/07 — Responding to PR feedback |
| A non-trivial architectural decision | examples/08 — Writing an ADR |
| A multi-week design needing critique | examples/09 — Writing an RFC |
| A web change ready to ship | examples/10 — Web QA before shipping |
| Multiple teammates' PRs to review (or any N independent tasks) | examples/11 — Multi-PR review with /orc:fan-out |
| A Jira ticket to link to a session and close on PR merge | examples/12 — Linking a Jira ticket and shipping with Resolves <KEY> |
Each example follows the same shape — Scenario → Flow → Walk-through → Artifacts → Done when → Variants → Iron rules in play — so you can scan to the relevant section.
orc is published as a single-plugin marketplace at this repo. Inside Claude Code:
/plugin marketplace add HigorAlves/orc
/plugin install orc@orc
The first command registers https://github.com/HigorAlves/orc as a marketplace named orc; the second installs the orc plugin from it. Updates pull with /plugin update orc@orc.
To pin a specific commit/tag, use the longhand source form in ~/.claude/settings.json:
npx claudepluginhub higoralves/orc --plugin orcStreamline your git workflow with simple commands for committing, pushing, and creating pull requests
Frontend design skill for UI/UX implementation
Analyze codebases to discover issues and create GitHub issues
Adds educational insights about implementation choices and codebase patterns (mimics the deprecated Explanatory output style)
Interactive learning mode that requests meaningful code contributions at decision points (mimics the unshipped Learning output style)
Harness-native ECC operator layer - 67 agents, 278 skills, 94 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
v9.52.0 - Reliability wave: tangle contextual review correction loop with hard round ceiling, progress-supervised review rounds (per-agent stall watch, descendant-tree kills), council diversity and agy pin fixes, marketplace generator source-of-truth fix, provider troubleshooting runbook and cost-expectations docs. Run /octo:setup.
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
Claude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows