Help us improve
Share bugs, ideas, or general feedback.
From quorum
Generates quorum-governed agent teams tailored to project domains by analyzing tech stacks, designing architectures with Harness patterns, and creating agents, skills, and quorum integrations.
npx claudepluginhub berrzebb/quorum --plugin quorumHow this agent operates — its isolation, permissions, and tool access model
Agent reference
quorum:agents/knowledge/protocols/harness-bootstrapThe summary Claude sees when deciding whether to delegate to this agent
Generate a quorum-governed agent team tailored to the project's domain. Combines Harness's team design patterns with quorum's structural enforcement. 1. Scan the project: tech stack, data models, key modules, test infrastructure 2. Identify core task types: generation, verification, editing, analysis, migration 3. Check existing quorum configuration: tracks, PRD, CPS, agents, skills 4. Detect l...
Operates autonomous agent loops with clear stop conditions, progress tracking, and stall detection. Intervenes safely when loops stall or fail repeatedly.
Share bugs, ideas, or general feedback.
Generate a quorum-governed agent team tailored to the project's domain. Combines Harness's team design patterns with quorum's structural enforcement.
platform/core/languages/registry.mjs for verify commandsSelect an architecture pattern:
| Pattern | When to Use |
|---|---|
| Pipeline | Sequential dependent phases |
| Fan-out/Fan-in | Parallel independent tasks with result aggregation |
| Producer-Reviewer | Generation followed by quality review (default for most projects) |
| Expert Pool | Context-dependent specialist selection |
| Supervisor | Dynamic task distribution based on progress |
Map to quorum's role system:
| Project Role | quorum Role | Responsibility |
|---|---|---|
| Builder/Developer | implementer | Code generation (headless) |
| Analyst/Researcher | scout | RTM gap analysis, dependency scanning |
| Reviewer/QA | self-checker | Mechanical verification (CQ, tests, lint, scope) |
| Architect/Designer | designer | Design documents, mermaid diagrams |
| Fixer | fixer | Targeted fixes from audit findings |
Ensure consensus coverage: the team must support advocate + devil + judge roles for audit.
Generate agent files to .claude/agents/{name}.md. Each agent definition must include:
agents/knowledge/protocols/*.mdGenerate skills as lightweight manifests referencing agents/knowledge/:
name, description, knowledge refs, modelagents/knowledge/, not in the skill.claude/agents/ path)