opencell-tools
A collection of Claude Code plugins that streamline the OpenCell developer workflow — from Jira ticket management and UI scaffolding to automated commits, pull requests, and code review.
Installation
- Add the marketplace:
claude
/install-plugin marketplace https://bitbucket.org/opencellsoft/ccode-marketplace.git
- Install the plugins you need — every plugin in the tables below installs the same way, with
/plugin install <plugin>@opencell-tools. For example, the frontend set:
/plugin install oc-fe-engineer@opencell-tools
/plugin install oc-fe-designer@opencell-tools
/plugin install oc-fe-reviewer@opencell-tools
/plugin install oc-fe-test-writer@opencell-tools
/plugin install oc-fe-create-ui@opencell-tools
/plugin install oc-fe-fix-bug@opencell-tools
/plugin install oc-fe-fix-pr@opencell-tools
Plugins
Skills (slash commands)
| Plugin | Command | Description |
|---|
| oc-cache-jira | /oc-cache-jira | Fetch and cache Jira ticket data locally for use by other commands |
| oc-commit | /oc-commit | Commit changes using the cached Jira ticket ID and summary, with automatic code review |
| oc-pull-request | /oc-pull-request | Push changes and create a pull request for the current Jira ticket (auto-detects Bitbucket vs GitHub) |
| oc-review-pr | /oc-review-pr | Review a pull request linked to a Jira ticket — fetch the PR, select the appropriate reviewer agent, and generate a detailed report |
| oc-fe-create-ui | /oc-fe-create-ui | Scaffold a UI page from a Jira ticket using the oc-fe-engineer sub-agent |
| oc-fe-fix-bug | /oc-fe-fix-bug | Fix a bug from a Jira ticket — update status to In Progress, create a fix branch, and start fixing |
| oc-fe-fix-pr | /oc-fe-fix-pr | Fix a pull request's review remarks — read its unresolved Bitbucket comments, fix them on the PR branch via oc-fe-engineer, write tests, then reply, resolve, and push |
| oc-fe-test-writer | /oc-fe-write-tests | Write Vitest tests for changed code (git diff) or specific files via the oc-fe-test-writer sub-agent |
| oc-fe-create-e2e-test | /oc-fe-create-e2e-test | Create Playwright E2E tests from Jira ticket requirements via the oc-fe-e2e-expert sub-agent |
| oc-fe-regression-test | /oc-fe-regression-test | Diff-driven Playwright regression specs for changed OpenCell Portal screens, written on the current branch right after the Vitest step so they ship in the PR |
| oc-ar-tools | /oc-ar-tech-design | Analyze a user story and produce a technical design for Opencell Core |
| oc-ar-ai-tools | /oc-ar-ai-tech-design | Technical design tailored for AI-assisted (Claude) development |
Backend toolkit (oc-be-tools)
oc-be-tools is a complete backend development toolkit for Opencell Core. It bundles slash commands, builder/reviewer sub-agents, and guideline-loading skills that all share one source of truth: the guideline files under plugins/backend/oc-be-tools/guidelines/ (CRITICAL_RULES, ENTITY, SERVICE, API, DATABASE, CODE_QUALITY, TESTING).
| Type | Provides | Description |
|---|
| Commands | /oc-be-implement | Orchestrate the full implementation of a Jira ticket across entities, database, services, API, tests, and Postman collections |
| Commands | /oc-be-review | Evaluate backend changes against the guidelines — reviews uncommitted code or a specific Bitbucket PR (with approve/request-changes) |
| Sub-agents | oc-be-entity-builder, oc-be-service-builder, oc-be-api-builder, oc-be-test-generator, oc-be-postman-generator | Builder agents that scaffold each layer following the guidelines |
| Sub-agents | oc-be-pr-reviewer | Backend code reviewer validating Java/EJB/JPA/Liquibase code against the guidelines, with a score and file:line suggestions |
| Skills | /oc-be-entity-guide, /oc-be-service-guide, /oc-be-api-guide, /oc-be-db-guide | Skills that load the relevant guidelines when working on each layer |
Overlay toolkit (oc-ov-tools)
oc-ov-tools covers Opencell overlay repositories — projects such as opencell-vertical-energy
and overlay-project-template that build the final opencell.war by layering their own jars and
resources over core Opencell's war.
It is a delta layer: the oc-be-tools guidelines stay authoritative and load first, and the
overlay files state only what differs. Install oc-be-tools too — oc-ov-tools hard-stops
without it. No configuration is required; the core guidelines are located automatically.
/plugin install oc-be-tools@opencell-tools
/plugin install oc-ov-tools@opencell-tools
| Type | Provides | Description |
|---|
| Commands | /oc-ov-implement | Orchestrate a full overlay ticket, with core-repo, branch-target and Liquibase gates |
| Commands | /oc-ov-review | Guideline + requirements + Sonar review, plus nine overlay mechanical gates |
| Commands | /oc-ov-calculate-ai-use | Thin alias over the shared backend AI-usage command |
| Sub-agents | oc-ov-entity-builder, oc-ov-service-builder, oc-ov-api-builder, oc-ov-script-builder | Entities + overlay.xml pair, @Specializes core-service extension, apiv0 REST, ScriptInstances and Jobs |
| Sub-agents | oc-ov-pr-reviewer | Overlay-aware reviewer — no AGPL / Immutables / activator false positives |
| Skills | /oc-ov-entity-guide, /oc-ov-service-guide, /oc-ov-api-guide, /oc-ov-db-guide, /oc-ov-script-guide, /oc-ov-test-guide, /oc-ov-postman-guide, /oc-ov-jasper-guide, /oc-ov-override-guide | Load the core guidelines, then the matching overlay delta |
Why the layer exists: an overlay uses no AGPL header, has no JAX-RS activator (core discovers apiv0
resources by Reflections scan), uses plain-POJO DTOs rather than Immutables, keeps Liquibase in
overlay.xml rather than structure.xml, and cannot use @Inject inside ScriptInstances. It also
adds a whole subject core has no equivalent for: overriding core, preferably via CDI @Specializes.
Functional toolkit (oc-fn-tools)
oc-fn-tools (func factory) bundles the functional / product-design skills for Opencell. These are auto-loading skills — no slash command; each triggers from context — covering Jira authoring, Confluence docs, portal capture, and the design-first delivery methodology.
| Type | Provides | Description |
|---|
| Skills | oc-fn-func-design | Author Jira INTRD issues (Epic / User Story / Enabler / Bug / Feature) — templates, ADF custom fields, acceptance criteria. Functional lane; defers Technical-design authoring to oc-ar-tech-design. |
| Skills | oc-fn-documentation | Create / update Confluence pages in the Opencell docs space (Concepts + User Manuals) |
| Skills | oc-fn-portal | Drive the Opencell Portal via Playwright for design/docs screenshots (not testing); ships the headless oc-fn-playwright MCP server |
| Skills | oc-fn-project-management | Design-first phased delivery methodology — phase gates, ADRs, repo/CI conventions |
Requires an Atlassian Rovo MCP connection for the Jira/Confluence skills — see
Atlassian and Bitbucket access.
Sub-agents (frontend)
| Plugin | Description |
|---|
| oc-fe-engineer | Expert React/TypeScript sub-agent for building, refactoring, and architecting frontend components |
| oc-fe-designer | Frontend designer sub-agent for translating Figma designs into implementation-ready React/MUI components with design tokens and styling |
| oc-fe-reviewer | Frontend code reviewer sub-agent that validates React/TypeScript code against project standards |
| oc-fe-test-writer | Frontend test writer sub-agent that writes Vitest tests for changed React/TypeScript code |
| oc-fe-cypress-expert | Cypress testing expert sub-agent for end-to-end testing, test automation, and flaky test resolution |
| oc-fe-e2e-expert | Playwright E2E expert sub-agent for authoring browser end-to-end tests |
MCP Integrations
| Plugin | Command | Description |
|---|
| oc-figma-mcp | /oc-figma | Extract design context, generate code from Figma designs, and retrieve design tokens |
| oc-playwright-mcp | /oc-playwright | Automate browser interactions, take screenshots, and test web applications |
| oc-opencell-mcp | /oc-opencell | Query and manage the Opencell billing system — invoices, quotes, customers, payments, subscriptions, and more |
| oc-sonar-mcp | — | Access SonarQube code quality metrics, issues, and analysis results |
| oc-postgres-mcp | — | Database health analysis, index tuning, query optimization, and safe SQL execution via PostgreSQL |
Atlassian and Bitbucket access
The Jira-driven workflow (/oc-cache-jira → /oc-commit → /oc-pull-request → /oc-review-pr →
/oc-fe-fix-pr) touches two systems, reached two different ways.
Jira / Confluence — official Atlassian plugin, no credentials. Install it from Anthropic's
marketplace (available by default) and sign in once:
/plugin install atlassian@claude-plugins-official
/mcp # complete the browser sign-in for the `atlassian` server
Bitbucket — email + API token. The Rovo MCP server exposes its Bitbucket tools only under
API-token auth, never over the OAuth flow the official plugin uses, so an OAuth Rovo connection
provides no bitbucket* tools. Every Bitbucket operation therefore goes through the REST API. Create
an Atlassian API token at https://id.atlassian.com/manage/api-tokens and export both values:
export BITBUCKET_EMAIL="[email protected]"
export BITBUCKET_ACCESS_TOKEN="your-api-token" # ATATT…
Both are required: an Atlassian API token authenticates with Basic auth as email:token
(curl -u "$BITBUCKET_EMAIL:$BITBUCKET_ACCESS_TOKEN"). Sending it as Authorization: Bearer returns
401. A Bitbucket repository/workspace Access Token is the other valid credential type and does
use Bearer without an email. App Passwords were removed on 2026-07-28.
Without working credentials, /oc-pull-request falls back to printing a PR-creation URL, and
/oc-review-pr / /oc-fe-fix-pr cannot reach the PR.
