Now in private beta — Claude Code, GitHub Copilot, Gemini & more Request access →
v0.4 · go 1.24+ · Private Beta

Stop winging it.
Ship features with a system.

bc-agentic-os gives your AI agents a shared brain — structured roles, stage-gate approvals, and persistent memory — so you build faster without breaking things.

CLI + MCP server 11-stage workflow Works with your agent
~/projects/acme-apizsh
Works with
The problem

AI agents are powerful.
But chaotic.

Every session starts from scratch. Every suggestion bypasses review. You're the glue — the reviewer, the planner, the memory. bc-agentic-os replaces that with structure.

× Without a system
  • Agents jump straight to code, skipping design.
  • No shared memory between sessions.
  • Approvals? Non-existent. Mistakes? Costly.
  • You're the only source of truth — and you're tired.
With bc-agentic-os
  • Stage-gated workflow: explore → PRD → design → plan → verify.
  • Persistent memory via DevMemory MCP.
  • SDD gate blocks execution until all stages are approved.
  • Role-based agents: CPO, Architect, Dev, QA — each focused, isolated.
How it works

From idea to code.
One stage at a time.

Eleven stages. Each owned by a specialized agent. Each gated until approved. No more half-baked features shipping on vibes.

01
Explore
Research Agent
Discover requirements, risks, and constraints.
02
PRD
CPO Agent
Define the product requirements document.
03
Spec
CPO + Architect
Nail technical specs and API contracts.
04
Design
Architect Agent
System design, DB schema, architecture decisions.
05
Tasks
Orchestrator
Break work into executable, testable tasks.
06
Plan
Dev Agent
Implementation plan with file-level detail.
07
Verify
QA Agent
Review, test, and confirm it's shippable.
SDD Gate active · branch & exec locked until all stages approved run_a1b2c3 · feat/add-stripe-payments
run_a1b2c3bc-agentic
$ bc-agentic next run_a1b2c3 # → Loads PRD template, hands off to CPO agent $ bc-agentic approve run_a1b2c3 prd Stage approved. Unlocking: spec $ bc-agentic branch run_a1b2c3 feat/run_a1b2c3 created (SDD gate passed ✔)

The full pipeline — what actually happens

Each stage has a defined owner, a concrete output, and a human approval gate. Nothing moves forward without sign-off.

01
🔬 Explore bc-research

Maps the problem space. Identifies risks, constraints, and unknowns before any code is written. Everything that could go wrong is surfaced here.

Output → Explore Report
02
📋 PRD bc-cpo

Translates research into a structured product document. Defines scope, success criteria, edge cases, and user flows.

Output → Product Requirements Document
03
📐 Spec bc-cpo + bc-architect

Bridges the PRD and implementation. Defines API contracts, data models, integration points, and non-functional requirements.

Output → Technical Spec
04
conditional
🎨 Design Review bc-designer

Evaluates UI/UX impact before architecture commits. Produces component proposals, accessibility requirements, and design constraints.

Output → Component Proposals
05
🏛️ Design bc-architect

Produces the SDD — the gate document. Database schema, service boundaries, architecture decisions, and trade-offs. Nothing executes until this is approved.

Output → System Design Document (SDD)
06
⚡ Tasks bc-orchestrator

Breaks the SDD into discrete, testable work units. Each task maps to a specific agent role with clear scope and acceptance criteria.

Output → Task Manifest
07
📝 Plan bc-*-dev

File-level implementation intent. No code yet — pure planning. Prevents surprises, ensures alignment before execution starts.

Output → Implementation Plan
08
⚙️ Execute bc-*-dev

Dev agent implements the plan against the approved SDD. The branch is only created after all upstream stages are approved — the SDD gate enforces this.

Output → Working Code
09
conditional
✦ Design Validation bc-designer

Designer reviews the built UI against the original spec. Catches drift between design intent and implementation before QA.

Output → Compliance Report
10
✅ Verify bc-qa

QA agent validates correctness, runs audits, checks coverage, and confirms the implementation is shippable against spec.

Output → QA Report
11
📊 Report bc-orchestrator

Orchestrator summarizes the run, updates persistent memory with decisions and lessons, and closes the loop.

Output → Run Summary
Specialized agents

14 agents. One per role.
Zero confusion.

Every agent has a single responsibility and deep domain knowledge for its stage. They don't step on each other — they hand off cleanly.

bc-orchestrator
Lead SWE coordinator
Routes work to specialized agents, enforces stage gates, manages handoffs, handles incidents and corrective loops.
📋
bc-cpo
Product strategist
Shapes ideas into PRDs, defines scope, prioritizes features, makes business and user-facing decisions.
🎨
bc-designer
Design system specialist
Proposes UI components during Design Review, validates implementation compliance during Design Validation, enforces accessibility standards.
🏛️
bc-architect
System designer
Produces System Design Documents (SDDs), designs APIs, plans infrastructure, evaluates technical trade-offs.
⚛️
bc-next-dev
Next.js specialist
Implements Next.js/React apps with App Router, Server Components, Server Actions, Route Handlers, and TailwindCSS.
📱
bc-flutter-dev
Flutter specialist
Implements Flutter/Dart apps (mobile, web, desktop). Handles state management, routing, REST integration.
🌐
bc-dotnet-dev
.NET API specialist
Builds production-grade ASP.NET Core APIs with EF Core, PostgreSQL, JWT auth, and clean architecture patterns.
🗄️
bc-db-admin
Database specialist
PostgreSQL schema design, EF Core migrations, query optimization, index strategy, and DB incident resolution.
🚀
bc-devops
CI/CD & deployment
CI/CD pipelines, Docker builds, GitHub Actions workflows, and deployment configuration for bc-agentic projects.
🔒
bc-security
Security specialist
OWASP security reviews, dependency audits, auth pattern validation, and secret scanning across bc-agentic projects.
🎬
bc-remotion
Video generation specialist
Creates programmatic videos using Remotion. Generates dynamic video content from code and data.
🎞️
bc-short-script-planner
Short-form content planner
Generates structured scripts for programming shorts — YouTube Shorts, TikTok, and Reels — with synchronized subtitles and code walkthroughs.
🔬
bc-research
Discovery specialist
Explores codebases, libraries, and documentation. Provides technical research and feasibility analysis.
bc-qa
Quality assurance
Verifies implementation, runs audits, checks accessibility, validates against spec and SDD.
Execution modes

You control how much the AI does.

From paranoid to full throttle. Every run declares its mode. Every mode has a guardrail.

safe · default
Read-only agent
Agent suggests edits, shell calls, and changes. You approve every one.
apply
Guarded execution
Agent writes files & runs whitelisted shell. You review the diff before commit.
god
Skip manual approvals
Stage gates still enforced. Approvals auto-granted inside the run. Maximum velocity.
full
Unrestricted
No guardrails. Full autonomy. Use on throwaway branches & sandboxes only.
Capabilities

Built for real development workflows.

Not a prompt wrapper. A full operational layer for multi-agent development — opinionated where it matters, open everywhere else.

14 specialized agents
Orchestrator, CPO, Architect, Designer, Next.js Dev, Flutter Dev, .NET Dev, DB Admin, DevOps, Security, Research, Remotion, Short-Script, QA — each isolated and focused.
SDD Gate
branch and exec are blocked until every stage is approved. No shortcuts, no surprises.
Persistent memory
DevMemory MCP keeps run state, decisions, and summaries across sessions and across clients — nothing is lost between context windows.
Multi-client support
One install. Runs across Claude Code, GitHub Copilot, OpenCode, and Gemini CLI — same workflow, same memory, any client.
Context7 integration
Accurate, up-to-date library docs pulled on demand. No hallucinated APIs, no outdated examples, no guessing.
Stack-aware routing
The orchestrator routes work to the right dev agent automatically. Flutter task goes to bc-flutter-dev. Next.js task goes to bc-next-dev. No configuration needed.
Human-in-the-loop
Safe mode by default. You review and approve at every stage gate. God mode skips approvals but never skips the SDD gate. You always stay in control.
Production-grade artifacts
Every stage produces a reusable document: PRDs, SDDs, specs, task manifests, test reports — all stored in docs/agentic and tracked in memory.
Cross-platform binary
Go binary. Runs on Linux, macOS, and Windows. Zero runtime dependencies. Dual-mode: CLI for interactive runs or MCP server for IDE integration.
Get access

Ready to bring structure
to your AI workflows?

bc-agentic-os is a commercial product in private beta. Reach out to discuss access, pricing, and onboarding.

What you get
  • Full bc-agentic-os binary (CLI + MCP)
  • All specialized agents provisioned
  • DevMemory MCP for persistent state
  • Setup & onboarding support
  • Private beta access with direct feedback channel
Reach out directly
boricua.coder@gmail.com
Private beta — limited seats Response within 24h