By FroncekB
Pell Software's Claude Code toolkit: PR/local code reviewers (correctness, quality, security), Jira workflow ops, and house-style guidance. One plugin, many skills.
Pull the inline + general comments off one of your Bitbucket PRs and triage each — apply mechanical fixes to your working tree or reply on the thread. Never commits, never pushes, never resolves threads.
Review a code change for correctness only (logic errors, broken invariants, missing error handling). Works on a Bitbucket PR or local changes. Returns a markdown report with all findings.
Close out a branch by opening a Bitbucket PR and (only on explicit consent) transitioning the linked Jira ticket and adding a PR-link comment. Read-only against Jira by default; PR creation always confirmed.
Compose ticket-to-implementation in one command. Fetches a Jira ticket, dispatches /pell:start-work to create a branch, then hands off to superpowers:brainstorming → writing-plans for the design and plan. When superpowers isn't installed, runs a lightweight inline substitute that produces a starter spec.
Three-pass review of local uncommitted changes — dispatches correctness, quality, and security reviewers that respect CLAUDE.md and surrounding code conventions; add "with tests" for an optional fourth test-coverage pass. Optionally applies suggested fixes.
Reviews a code change for correctness — logic errors, off-by-one bugs, broken invariants, missing error handling at real boundaries, race conditions, regressions, and mismatches with stated intent (Jira ticket or CLAUDE.md). Returns ALL findings including nits. Use as part of /pell:correctness-review, /pell:three-pass-review, or /pell:local-review.
Reviews a code change for code quality — readability, naming, duplication, premature abstraction, dead code, and especially adherence to repo conventions (CLAUDE.md, linter configs, surrounding patterns). Returns ALL findings including nits. Use as part of /pell:quality-review, /pell:three-pass-review, or /pell:local-review.
Reviews a repository (not a diff) for code quality, reusability opportunities, dead code, convention drift, and files that have grown too big. Operates on a focused list of files supplied by the dispatching command. Returns ALL findings including nits. Use as part of /pell:repo-review.
Reviews a repository (not a diff) for sensitive data (SSNs, credit cards, API keys, JWTs, private keys, driver's license patterns) and code-level vulnerabilities (XSS, SQLi, path traversal, hardcoded credentials, crypto misuse, PII logging). Operates on a focused list of files supplied by the dispatching command. Returns ALL findings including low-severity observations. Use as part of /pell:repo-security-review.
Reviews a code change for security vulnerabilities — injection, authn/authz gaps, secret leakage, unsafe deserialization, missing input validation at trust boundaries, OWASP top-10. Returns ALL findings including low-severity nits. Use as part of /pell:security-review, /pell:three-pass-review, or /pell:local-review.
Use at the START of any user-facing UI work in a Pell repo — building React components, pages, layouts, design systems, styling, Tailwind/CSS, marketing pages, dashboards, forms, or any task whose deliverable is something a human will look at in a browser.
Use when about to explain something inherently visual — architecture or data-flow diagrams, state machines, geometry or layout, before/after comparisons, tree or graph structures, timelines, or table-heavy comparisons — and a rendered view in a browser would communicate it better than terminal prose. Also use when soliciting a choice that's clearer as clickable options than as text.
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.
A Claude Code toolkit built around how Pell Software actually ships code.
This is an internal marketplace for skills, commands, and agents that codify the workflows Pell engineers use every day — reviewing pull requests, working tickets through GitFlow, keeping code consistent with our conventions. Install once, get the whole kit.
Three goals drive everything in this repo:
/plugin marketplace add FroncekB/pell_skills
/plugin install pell@pell-skills
/reload-plugins
That's it for the toolkit itself. Some commands need additional MCP servers configured — see the next section.
Most commands need at least one of these two Atlassian MCP servers connected to your Claude Code session.
| Server | Auth | Required for |
|---|---|---|
atlassian-bitbucket | API token (Basic auth) | Any PR-mode review (/pell:correctness-review, /pell:quality-review, /pell:security-review against a PR), /pell:three-pass-review, /pell:address-review, future Bitbucket-aware commands |
plugin:atlassian:atlassian | OAuth | Jira ticket context in /pell:three-pass-review, future Jira-ops commands |
The Atlassian Rovo MCP server exposes different tools depending on the auth method, and Claude Code dedups MCP server connections by URL. This means you need two parallel connections to the same backing server, at distinct endpoints. Here's the full setup:
1. Install the Atlassian OAuth plugin (for Jira/Confluence):
/plugin install atlassian@claude-plugins-official
It auto-prompts for OAuth on first use. Sign in with your Pell Atlassian account.
2. Create a scoped Bitbucket API token at id.atlassian.com/manage-profile/security/api-tokens. Required scopes:
read:user:bitbucketread:workspace:bitbucketread:repository:bitbucketread:pullrequest:bitbucketwrite:pullrequest:bitbucket (only if you want comment posting)3. Base64-encode your credentials:
echo -n "[email protected]:YOUR_API_TOKEN" | base64
4. Register the Bitbucket connection at the authv2 endpoint (this is the key step — using a distinct URL from the OAuth connection so they don't dedup):
claude mcp add-json atlassian-bitbucket '{
"type": "http",
"url": "https://mcp.atlassian.com/v1/mcp/authv2",
"headers": {
"Authorization": "Basic <YOUR_BASE64_STRING>"
}
}'
5. Restart Claude Code so both MCPs come up.
6. Verify both are connected with /mcp — you should see atlassian-bitbucket: ✓ Connected and the Atlassian plugin's tools available.
If something fails, see the troubleshooting note in docs/specs/2026-05-27-pell-skills-architecture.md §5 or check that the API token has all the scopes listed above.
/pell:correctness-reviewSingle-dimension review for correctness only. Surfaces logic errors, off-by-one bugs, broken invariants, missing error handling at real boundaries, race conditions, regressions, and CLAUDE.md violations.
Usage:
/pell:correctness-review # reviews `git diff HEAD`
/pell:correctness-review --staged # reviews staged changes only
/pell:correctness-review src/api/handlers/ # reviews changes under a path
/pell:correctness-review vrs_default#42 # reviews Bitbucket PR
/pell:correctness-review https://bitbucket.org/.../pull-requests/42
/pell:correctness-review 42 use bitbucket # PR mode with remote context fetch
Output: markdown report grouped by severity (blocker / major / minor / nit). Read-only — never modifies anything.
/pell:quality-reviewSingle-dimension review for code quality. Looks at readability, naming, duplication, dead code, premature abstraction, and especially adherence to repo conventions (CLAUDE.md, .editorconfig, .csharpierrc, surrounding code patterns).
Usage: same shape as /pell:correctness-review.
Output: markdown report grouped by severity (major / minor / nit). Read-only.
/pell:security-reviewnpx claudepluginhub froncekb/pell_skills --plugin pellHarness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.