By 99x
Requirement grooming plugin focused on user experience. Analyzes user intent, domain knowledge, competitive context, and workflow to produce well-understood, groomed requirements.
Domain and market context analyst. Brings general domain knowledge, competitive insights, data meaning (semantics not structure), and business rules from a user lens. Uses web search for industry patterns.
Gap, risk, and value analyst. Identifies ambiguities, failure modes, value/priority, and dependencies — grounded in intent and domain analyst outputs.
Intent and user-context analyst. Decomposes stated need into underlying intent, success definition, user context, and workflow mapping.
Requirement grooming orchestrator. Analyzes user intent, domain knowledge, competitive context, and workflow to produce a well-understood, groomed requirement focused on user experience.
Groom a GitHub issue. Analyzes user intent, domain knowledge, competitive context, and workflow. Usage: /analyze-requirement [issue-number]
Post the elaborated requirement to a GitHub issue. Requires an issue number. Usage: /post-elaboration [issue-number]
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.
Humans and AI agents, working as one team across the full software development lifecycle.
Xianix Team embeds a coordinated mesh of AI agents into every phase of the SDLC — from requirement analysis and sprint planning through to PR review, test strategy, and documentation maintenance. The goal is amplification, not replacement: human engineers operate at 10x efficacy while agents handle the repetitive, detail-heavy work that keeps quality and standards from slipping.
The first agent shipped is the PR Review Agent — a fully autonomous code reviewer that triggers on pull requests, analyses the diff against architecture rules and coding standards, and posts structured feedback directly on the PR.
| Document | Description |
|---|---|
| docs/concept.md | Vision, the full SDLC agent pipeline, and the agent mesh model |
| docs/agent-architecture.md | Technical architecture: Agent Control Plane, webhook flow, Claude Code plugin system |
| docs/webhook-provider-design.md | Webhook parsing design — provider identification, unified PR context model, GitHub vs Azure DevOps |
| Document | Description |
|---|---|
| docs/manual-plugin.setup.md | Install and use the PR Review plugin manually via Claude Code |
| plugins/pr-reviewer/docs/platform-setup.md | Platform setup for GitHub and Azure DevOps (review posting, tokens) |
| plugins/pr-reviewer/docs/git-auth.md | Runtime git credentials — how tokens are passed for clone and push |
| plugins/pr-reviewer/docs/mcp-config.md | MCP configuration reference (superseded by platform-setup.md) |
| Document | Description |
|---|---|
| docs/docker-deployment.md | Running the PR Review agent in Docker (build, run, volumes, CI/CD publish) |
| docs/dockerhub-publishing.md | Publishing the Docker image to Docker Hub via GitHub Actions |
The PR Review Agent connects to the Xians Agent Control Plane (ACP), which receives GitHub / Azure DevOps webhooks and relays them to the locally-running console process — no inbound port or public IP required.
GitHub / ADO
│ PR opened or synchronized → webhook POST
▼
Xians Agent Control Plane (ACP)
│ cloud relay — routes event to tenant's registered agent
▼
AgentTeam.Console (Program.cs — .NET 9)
│ connects to ACP via XIANS_SERVER_URL + XIANS_API_KEY
│ PrReviewAgent.Register() → DefineIntegrator().OnWebhook()
▼
PrReviewAgent (webhook handler)
│ WebhookParserResolver → GitHubWebhookParser | AzureDevOpsWebhookParser
│ extracts: PlatformName, RepoUrl, PrNumber, SourceBranch
│ builds deterministic workflowId (deduplicates re-delivered webhooks)
▼
PrReviewScriptWorkflow (Temporal — durable, resumable)
│ ActivityOptions: 20 min timeout, MaxAttempts = 1 (script is not idempotent)
▼
RunPrReviewScriptActivity
│ resolves scripts/run-pr-review.sh from XIANIX_REPO_ROOT or binary path
│ injects env vars: PLATFORM, REPO_URL, PR_NUMBER, PR_SOURCE_REF
│ inherits process env so tokens (GITHUB_TOKEN / AZURE_TOKEN) flow through
▼
bash scripts/run-pr-review.sh
│
├── git clone --bare (first run → REPO_CACHE_DIR, shared across concurrent PRs)
├── git fetch --prune (subsequent runs — incremental only)
├── git worktree add (isolated per-run checkout → WORKDIR)
├── git fetch refs/pull/<PR>/head
├── git clone xianix-team (first run → XIANIX_CACHE_DIR)
├── git pull xianix-team (subsequent runs — update plugin to latest)
├── write MCP config (token injected at runtime, never committed)
└── claude: /review-pr <PR_NUMBER> [--fix]
│
├── pr-reviewer agent
│ ├── code-reviewer
│ ├── security-reviewer
│ ├── test-reviewer
│ └── performance-reviewer
│
└── providers/github.md | providers/azure-devops.md
post review + inline comments
│
└── git worktree remove (cleanup — bare cache kept for next run)
Each PR review runs in its own git worktree — an independent checkout backed by a single shared bare clone. Concurrent reviews share the object store without conflict.
npx claudepluginhub 99x/xianix-team --plugin req-analyst-pluginComprehensive PR review plugin with specialized agents for code quality, security, test coverage, and performance analysis. Works with GitHub, Azure DevOps, Bitbucket, and any git repository.
Requirements-driven development workflow with quality gates for practical feature implementation
Autonomous development methodology: PRD interviews → agent execution → automated review
Complete PRP workflow automation - plan, implement, debug, review, commit, and PR commands with specialized agents
Implementation planning, execution, and PR creation workflows with multi-agent collaboration
Orchestration plugin. v1 use case: async development - turn ready issues into pull requests, then iterate on review feedback until a human takes over. Designed to host more orchestration use cases (refactoring, docs, audits) in future versions.
Helder's personal SDLC toolbelt for AI coding agents — from PRD to ship. Bundles the tracer-bullet workflow alongside TDD, code review, audits, and shipping skills.