bohewu/agents_pipeline
Codex-first multi-agent workflow toolkit with global skills, workspace-local model profiles, a runtime-neutral core, and Claude Code/Copilot exports.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code $ npx degit bohewu/agents_pipeline/agents ./rig-agents_pipeline/agents $ npx degit bohewu/agents_pipeline/skills ./rig-agents_pipeline/skills
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit bohewu/agents_pipeline/skills/artgen-scaffold .claude/skills/artgen-scaffold $ npx degit bohewu/agents_pipeline/skills/devtools-ux-audit .claude/skills/devtools-ux-audit $ npx degit bohewu/agents_pipeline/skills/frontend-aesthetic-director .claude/skills/frontend-aesthetic-director $ npx degit bohewu/agents_pipeline/skills/run-adaptive .claude/skills/run-adaptive $ npx degit bohewu/agents_pipeline/skills/run-analysis .claude/skills/run-analysis $ npx degit bohewu/agents_pipeline/skills/run-ci .claude/skills/run-ci $ npx degit bohewu/agents_pipeline/skills/run-committee .claude/skills/run-committee $ npx degit bohewu/agents_pipeline/skills/run-flow .claude/skills/run-flow $ npx degit bohewu/agents_pipeline/skills/run-general .claude/skills/run-general $ npx degit bohewu/agents_pipeline/skills/run-modernize .claude/skills/run-modernize $ npx degit bohewu/agents_pipeline/skills/run-pipeline .claude/skills/run-pipeline $ npx degit bohewu/agents_pipeline/skills/run-simple .claude/skills/run-simple $ npx degit bohewu/agents_pipeline/skills/run-spec .claude/skills/run-spec $ npx degit bohewu/agents_pipeline/skills/run-ux .claude/skills/run-ux $ npx degit bohewu/agents_pipeline/skills/ui-communication-designer .claude/skills/ui-communication-designer $ npx degit bohewu/agents_pipeline/skills/ui-ux-workflow .claude/skills/ui-ux-workflow
$ curl -fsSL --create-dirs -o .claude/agents/analysis-complexity.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/analysis-complexity.md $ curl -fsSL --create-dirs -o .claude/agents/analysis-correctness.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/analysis-correctness.md $ curl -fsSL --create-dirs -o .claude/agents/analysis-numerics.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/analysis-numerics.md $ curl -fsSL --create-dirs -o .claude/agents/analysis-robustness.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/analysis-robustness.md $ curl -fsSL --create-dirs -o .claude/agents/art-director.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/art-director.md $ curl -fsSL --create-dirs -o .claude/agents/atomizer.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/atomizer.md $ curl -fsSL --create-dirs -o .claude/agents/committee-architect.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/committee-architect.md $ curl -fsSL --create-dirs -o .claude/agents/committee-judge.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/committee-judge.md $ curl -fsSL --create-dirs -o .claude/agents/committee-kiss.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/committee-kiss.md $ curl -fsSL --create-dirs -o .claude/agents/committee-product.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/committee-product.md $ curl -fsSL --create-dirs -o .claude/agents/committee-qa.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/committee-qa.md $ curl -fsSL --create-dirs -o .claude/agents/committee-security.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/committee-security.md $ curl -fsSL --create-dirs -o .claude/agents/compressor.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/compressor.md $ curl -fsSL --create-dirs -o .claude/agents/debugger.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/debugger.md $ curl -fsSL --create-dirs -o .claude/agents/doc-writer.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/doc-writer.md $ curl -fsSL --create-dirs -o .claude/agents/executor-strong.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/executor-strong.md $ curl -fsSL --create-dirs -o .claude/agents/executor.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/executor.md $ curl -fsSL --create-dirs -o .claude/agents/flow-splitter.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/flow-splitter.md $ curl -fsSL --create-dirs -o .claude/agents/generalist.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/generalist.md $ curl -fsSL --create-dirs -o .claude/agents/handoff-writer.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/handoff-writer.md $ curl -fsSL --create-dirs -o .claude/agents/kanban-manager.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/kanban-manager.md $ curl -fsSL --create-dirs -o .claude/agents/market-researcher.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/market-researcher.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-analysis.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-analysis.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-ci.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-ci.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-committee.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-committee.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-flow.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-flow.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-general.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-general.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-modernize.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-modernize.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-pipeline.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-pipeline.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-simple.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-simple.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-spec.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-spec.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator-ux.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/orchestrator-ux.md $ curl -fsSL --create-dirs -o .claude/agents/peon.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/peon.md $ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/planner.md $ curl -fsSL --create-dirs -o .claude/agents/repo-scout.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/repo-scout.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/router.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/router.md $ curl -fsSL --create-dirs -o .claude/agents/session-guide-writer.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/session-guide-writer.md $ curl -fsSL --create-dirs -o .claude/agents/specifier.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/specifier.md $ curl -fsSL --create-dirs -o .claude/agents/summarizer.md https://raw.githubusercontent.com/bohewu/agents_pipeline/main/agents/summarizer.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (16)
artgen-scaffolddevtools-ux-auditfrontend-aesthetic-directorrun-adaptiverun-analysisrun-cirun-committeerun-flowrun-generalrun-modernizerun-pipelinerun-simplerun-specrun-uxui-communication-designerui-ux-workflow
Subagents (47)
| analysis-complexity | Analysis expert focused on time/space complexity, scalability, data structure fitness, and algorithmic efficiency. |
| analysis-correctness | Analysis expert focused on logical correctness, invariants, state consistency, and proof-level reasoning. |
| analysis-numerics | Analysis expert focused on numerical stability, floating-point precision, accumulation errors, and mathematical correctness. |
| analysis-robustness | Analysis expert focused on edge cases, boundary conditions, error paths, adversarial inputs, and defensive coding gaps. |
| art-director | Converts raw 2D asset requests into provider-neutral prompts or complete versioned brief and handoff packages. |
| atomizer | Converts PlanOutline (+ optional RepoFindings) into atomic tasks with DoD and dependencies (DAG). |
| committee-architect | Committee expert focused on architecture, maintainability, and long-term ownership. |
| committee-judge | Final judge for committee outputs. Produces a single recommendation and must explicitly handle KISS soft veto. |
| committee-kiss | KISS/complexity guard for committee decisions. May raise a soft veto against unnecessary complexity. |
| committee-product | Committee expert focused on user impact, product value, and scope discipline. |
| committee-qa | Committee expert focused on testing strategy, reliability, observability, and regressions. |
| committee-security | Committee expert focused on security, privacy, compliance, and risk containment. |
| compressor | Compresses repo findings + decisions + task results into a small, reusable context pack. |
| debugger | On-demand diagnosis specialist for difficult, uncertain, cross-module, or conflicting-evidence failures. |
| doc-writer | Documentation specialist for design/spec/checklist/analysis outputs. |
| executor-strong | Execute one difficult atomic task with the shared executor contract on a profile-owned strong binding. |
| executor | Executes one atomic task with bounded risk, verification, repair, and resource controls supplied in the handoff. Must provide evidence. |
| flow-splitter | Converts a Flow ProblemSpec (+ optional RepoFindings) into a max-5 atomic task list with bounded routing hints. |
| generalist | General-purpose executor for mixed-scope tasks. |
| handoff-writer | Produces run-local handoff artifacts and optional publishable handoff output for a new session. |
| kanban-manager | Manages canonical todo-ledger.json data and renders kanban.md. |
| market-researcher | Research specialist for web-based market scans, comparable products, pricing signals, and monetization benchmarks. |
| orchestrator-analysis | Post-hoc analysis pipeline with conditional expert roster, parallel analytical memos, and severity-ranked findings report. |
| orchestrator-ci | CI/CD planning orchestrator with docs-first outputs and optional generation. |
| orchestrator-committee | Swarm committee orchestrator for decision-making with multiple expert memos, a KISS soft-veto, and a final judge. |
| orchestrator-flow | Flow Orchestrator with atomic tasks, bounded flow, bounded parallelism, and max-5 task limit. |
| orchestrator-general | General-purpose orchestrator for mixed repository work: coding, planning, analysis, writing, maintenance, debugging, and handoff recommendations. |
| orchestrator-modernize | Experimental modernize pipeline for legacy systems. Produces modernization strategy docs and can optionally transition the current/main agent into the Pipeline workflow for implementation, including b |
| orchestrator-pipeline | Primary orchestrator for the full pipeline with risk-derived execution rigor, review gates, retries, and context compression. |
| orchestrator-simple | Simple build-style dispatcher that decomposes a request, delegates to subagents, and returns a concise result without writing run manifests or pipeline artifacts. |
| orchestrator-spec | Thin docs-first orchestrator for producing review-ready development specs that humans can read and later pipelines can implement. |
| orchestrator-ux | UX audit orchestrator for profile-aware normal-user scoring, multi-perspective findings, and actionable reports. |
| peon | Low-cost executor for mechanical or repetitive tasks. |
| planner | Produces a high-level plan from ProblemSpec (milestones, dependencies, deliverables). |
| repo-scout | Repo scouting specialist. Locates files, patterns, architecture, and risks with minimal cost. |
| reviewer | Quality gate and ad hoc reviewer. Validates DoD when pipeline task contracts exist, otherwise reviews explicit targets without requiring pipeline artifacts. |
| router | Builds a dispatch plan: assigns tasks to agents, batching, and parallel lanes. |
| session-guide-writer | Creates or refreshes stable repo-level guidance in session-guide.md. |
| specifier | Extracts structured requirements into ProblemSpec JSON and optional DevSpec JSON. No solutions. |
| summarizer | Produces the final user-facing summary. |
Similar rigs
joaolozano-lendario/athena-os
A Claude Code Native System for Intelligent Workflow Architecture - Slash Commands for Smarter AI Workflows
Orchestrator 3.7k tok · · Q15
orynvn/claude-workspace-setup
A project template for Claude Code CLI — opinionated, token-efficient, and built for multi-agent workflows.
Orchestrator 22.3k tok · · Q18
Yassinekrn/claude-setup
Personal Claude Code setup — orchestrator/subagent delegation policy, custom agents, and a colored PowerShell statusline (git branch, model+effort, tokens, cost, rate limits).
Orchestrator 1.2k tok · · Q15
AkshatNaruka/aidentity
A verified index of 526 AI agent skills — licenses, sources, one-command install for Claude Code, Codex, Cursor, OpenCode, Copilot and Gemini CLI.
Orchestrator 2.6k tok · · Q5