By sujeet-pro
Self-contained, Claude-optimized engineering skills: implement, review, pr-review, investigate, document. Each skill ships its own phased workflow and tailored persona. GitHub via the gh CLI, git operations via git, SSH-only clones.
Findings-first code reviewer. Reviews a diff one dimension at a time (correctness → tests → security → performance → readability → consistency), tiers every finding by severity, and quotes file:line evidence. Read-only — never edits, posts, or merges. Spawn one per review dimension for parallel fan-out.
Parallel one-hop context fetcher. Given the URLs and IDs in a prompt (Jira / GitHub / Confluence / Slack / Datadog / GDoc), fetches each via the right tool and returns a single deduplicated context summary. One hop only — never follows links found inside fetched content. Quotes tightly, surfaces gaps honestly.
Reader-first documentation author. Produces markdown artifacts (runbook, ADR, RCA, PR body, commit message, changelog, diagram, README, migration guide, API reference). Leads with the reader's question, cites every non-trivial claim to a repo path or quoted source, and cuts filler. Does not publish.
Careful code mutator. Reads every file before writing it, applies the smallest correct change, matches repo conventions, validates at boundaries. Uses git directly for version control and the gh CLI for any GitHub interaction. Never force-pushes, never commits to a protected branch, never merges.
Multi-source incident investigator. Correlates at least two independent signals before naming a root cause, states confidence on every claim, pins an explicit time window on every query, and recommends the lowest-blast-radius next action. Read-only — never modifies a monitor, dashboard, flag, or experiment. Never names a person as root cause.
Draft a technical document from an intent or a source. Triggers on "write the runbook / ADR / RCA / PR description / commit message / changelog / diagram / README / migration guide / API reference / experiment report / incident summary / onboarding doc / design doc". Markdown-first: produces a draft in a local file and NEVER publishes (no Confluence / Jira / Slack / GitHub posting — that is a separate concern). Reader-first voice: leads with the reader's question, cites every non-trivial claim to a repo path or quoted source, caps external quotes at 15 words, cuts filler. Audience-tuned (engineer / pm / exec / mixed) — the voice does not mix. GitHub context (PR / issue) is read via the gh CLI.
Implement, build, ship, or wire up a change in an existing repo. Polymorphic on input: a Jira URL/KEY-123, a GitHub issue URL or #N (fetched via the gh CLI), a Confluence/Slack link (fetched via MCP), or freeform "build the X" prose. Plans before it acts: gathers context, presents 2-4 approaches with trade-offs, confirms, then writes the smallest correct change matching repo conventions, with tests for new behavior. Validates with the repo's own typecheck/lint/narrow tests at every checkpoint. Git via git directly (feature branch, never protected, never force, never merge); GitHub via the gh CLI only (PRs via gh pr create); clones SSH-only.
Investigate, diagnose, RCA, "why is X slow/broken/down", "what changed" for any production symptom. Triggers on a free-text symptom + service, a Datadog incident/monitor/dashboard/log URL, a Slack alert permalink, a Statsig gate/experiment URL, or a Mixpanel/Snowflake/Looker question. Read-only — never modifies a monitor, dashboard, flag, or experiment, and never triggers a rollback or restart (recommends; the human executes). Pins an explicit time window on every query, correlates >=2 independent signals before naming a root cause, and states confidence (low/med/high) with anchored evidence and <=15-word verbatim quotes. Produces a timeline + hypothesis + a lowest-blast-radius next action.
HEAVYWEIGHT, GitHub-only deep PR review. Takes a GitHub PR URL, prepares a read-only git worktree at the PR head (SSH clone only), and reviews the diff with full cross-file context via Read/Grep/Glob. Fans out one agent per dimension (correctness, security, tests, performance, api, docs, observability, concurrency, feature-flow) through the Workflow tool, then adversarially verifies every finding before it survives. Traces feature flags / experiments via the adk-statsig MCP and pulls linked Jira/Confluence via the adk-atlassian MCP. Posts inline review comments, a review summary, and appreciations through the gh CLI after confirmation. NEVER merges, never force-pushes, never edits the PR's code. For a quick read-only pass with no worktree, use /adk:review.
Review, audit, sanity-check, look-at any review-able target. Triggers on a GitHub PR URL (fetched via the gh CLI), a local path or "." (review the working tree), a markdown/doc file, or a comment-thread URL. Read-only by default; --fix applies accepted findings locally and pushes to the PR branch after confirmation (never force, never merge, never to a protected branch). Produces severity-tiered findings (blocker / critical / should / may / nit) with path:line and <=15-word verbatim evidence quotes. Six dimensions: correctness, tests, security, performance, readability, consistency. For a deep PR review with cross-file code-context retrieval, use /adk:pr-review.
External network access
Connects to servers outside your machine
Requires secrets
Needs API keys or credentials to function
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.
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
A focused, Claude-optimized engineering toolkit, packaged as a single-plugin Claude Code marketplace. Install it and you get five self-contained skills, a set of tailored sub-agents, and pre-wired (env-gated) MCP servers.
Every skill is self-contained (its whole contract lives in its own folder), ships its own phased workflow and tailored persona, and uses the Workflow tool to fan out the heavy steps (multi-dimension review, multi-source investigation) with adversarial verification.
In Claude Code:
/plugin marketplace add sujeet-pro/agents-devkit
/plugin install adk@agents-devkit
Then invoke any skill with its namespaced command:
/adk:review <pr-url | . | path | doc>
/adk:pr-review <github-pr-url>
/adk:implement <jira | gh-issue | "build the X">
/adk:investigate <symptom | datadog-url | slack-alert>
/adk:document <intent> --type <runbook|adr|rca|pr-body|...>
Skills also auto-trigger by intent (e.g. "review this PR", "why is checkout slow") — the namespaced command is just the explicit way.
| Skill | What it does |
|---|---|
review | Lightweight, polymorphic review of a PR / local diff / doc / thread. Six dimensions, severity-tiered findings with file:line evidence. Read-only by default; --fix applies + pushes after confirmation. |
pr-review | Heavyweight, GitHub-only deep PR review: a read-only git worktree at the PR head, cross-file context, a Workflow that fans out one agent per dimension + feature-flag tracing + adversarial verification, then posts inline comments via the gh CLI. Never merges. |
implement | Build a change from a Jira ticket / GitHub issue / Slack thread / freeform prose. Plans before it acts, writes the smallest correct change with tests, validates with the repo's own tooling. |
investigate | Read-only multi-source RCA: pins an explicit window, fans out one agent per data source (Datadog / Slack / Statsig / Mixpanel / Snowflake / Looker / deploys), requires ≥2 agreeing signals before naming a root cause. |
document | Draft any markdown artifact (runbook / ADR / RCA / PR body / migration guide / …), reader-first and cited. Drafts locally; never publishes. |
Tailored sub-agents the skills' workflows spawn: code-reviewer, security-auditor, test-engineer, implementer, investigator, doc-writer, context-gatherer. Each is a sharp, single-role worker — see plugins/adk/agents/.
gh CLI only (gh pr view/diff/review, gh api). Run gh auth login first.git) for commits, branches, worktrees.git clone [email protected]:owner/repo.git). No HTTPS clones.The plugin ships .mcp.json with these servers. Each is opt-in via environment variables — set the ones for the data sources you use; the rest stay dormant. None are required for the core code skills (review, pr-review, implement use gh/git).
| MCP server | Used by | Environment variables |
|---|---|---|
adk-datadog | investigate | DATADOG_API_KEY, DATADOG_APP_KEY |
adk-atlassian | implement, document, investigate, pr-review | ATLASSIAN_SITE, ATLASSIAN_USERNAME, ATLASSIAN_API_TOKEN (needs uvx) |
adk-slack | investigate, implement | SLACK_BOT_TOKEN, SLACK_USER_TOKEN (needs npx) |
adk-statsig | pr-review, investigate | STATSIG_CONSOLE_API_KEY |
adk-mixpanel | investigate | OAuth on first use — no env vars |
adk-snowflake | investigate | SNOWFLAKE_CONNECTION_NAME, SNOWFLAKE_ACCESS_TOKEN, SNOWFLAKE_SERVICE_CONFIG_FILE, SNOWFLAKE_HOME (needs uvx) |
adk-looker | investigate | LOOKER_BASE_URL, LOOKER_CLIENT_ID, LOOKER_CLIENT_SECRET (needs uvx) |
adk-google | document, investigate | GOOGLE_CLIENT_ID, GOOGLE_CLIENT_SECRET, USER_GOOGLE_EMAIL, GOOGLE_WORKSPACE_MCP_CREDENTIALS_DIR (needs uvx) |
Set these in your shell environment before launching Claude Code; .mcp.json expands ${VAR} at connection time. A server whose env vars are unset simply won't connect, and the skills degrade honestly (they mark the source as skipped).
gh (authenticated: gh auth login)git with an SSH key configured for GitHubuvx (from uv) for the Atlassian / Snowflake / Looker / Google MCP serversnpx (Node) for the Slack MCP servernpx claudepluginhub sujeet-pro/agents-devkit --plugin adkHarness-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
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive C4 architecture documentation workflow with bottom-up code analysis, component synthesis, container mapping, and context diagram generation
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
AI-powered development tools for code review, research, design, and workflow automation.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer