- Home
- Skills
- Workflow Automation
- Conductor
Works with the AI Agents you already use
Conductor
A workflow orchestrator that sequences specialized agent roles through a rigid product development lifecycle.
$10
Conductor
Example session with this skill installed
Run the product crew on the 'Vault' project. Goal: build the local-first vault creation flow. vision is in ./brief.txt. No UI needed. Hosting is local-first. Stop at the planner.
- Read your context and instructions
- Compiled the conductor
I have initialized the product crew sequence for
Vault. The process stopped after the senior planning phase as requested. I have coordinated the Business Analyst, Software Architect, and Senior Planner roles to produce the necessary documentation and task breakdown.
Phase Plan: Vault Local-First Creation
| phase_id | owner_role | writes | reads | status |
|---|---|---|---|---|
| ba | business-analyst | docs/requirements.md | vision_path | done |
| ux | ux-designer | docs/ux.md | docs/requirements.md | skipped: no UI |
| sw-arch | software-architect | docs/architecture.md | docs/requirements.md | done |
| infra | system-architect | docs/infrastructure.md | docs/architecture.md | done |
| plan | senior-planner | plans/tasks.md | docs/arch, docs/reqs | done |
| tdd | junior-tdd | src/* | plans/tasks.md | queued |
| qa | qa-engineer | qa/report.md | tdd artifacts | queued |
Handoff Index
docs/requirements.md: P0 stories for local vault initialization and master key setup.docs/architecture.md: SQLite with AES-256-GCM encryption strategy defined.docs/infrastructure.md: Local-only storage spec, no cloud persistence.plans/tasks.md: 4 discrete TDD tasks (Init DB, Key Gen, Vault Schema, Write Ops).
Stop Conditions Met
- Planner phase complete. Implementation spine (TDD/QA) is queued but not started.
- Conductor residual:
implemented: no.
Next steps
- Review the task estimates in
plans/tasks.md. - Increase
wip_limitif you want to run parallel TDD tasks for non-overlapping files. - Call the
junior-tddskill to begin the build spine.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Large language models often skip critical planning phases, jumping straight to implementation and creating "spaghetti" architectures or missing core requirements. Coordinating multiple specialized agent roles without a strict operational sequence leads to scope creep and inconsistent file handoffs.
What it does
- Sequences a product crew through a specific build slice using a rigid phase plan.
- Enforces gates between Business Analyst, UX, Architecture, Infra, and TDD roles.
- Generates a
phase-plan.mdscoreboard to track role status, artifact writes, and stop conditions. - Prevents the conductor from authoring application code, ensuring role separation and quality control.
- Validates handoff artifacts against a strict contract before allowing the next phase to trigger.
Frameworks & tools
Works with any agent capable of file system operations. Specifically designed to orchestrate siblings like business-analyst, ux-designer, software-architect, senior-planner, and junior-tdd.
Why this beats prompting it yourself
DIY prompting often fails to maintain state across complex multi-step workflows, leading to "hallucinated" progress. This skill provides a structured state machine that requires physical artifact verification at every gate, preventing the model from skipping boring but essential steps like requirements gathering or infrastructure planning.
Use cases
- Orchestrate a full product team from vision to task list without writing a single line of code.
- Ensure a Junior TDD agent only works on tasks pre-approved by a Senior Planner and Architect.
- Audit a development workflow by forcing agents to document handoffs in a standardized index.
- Stop a build immediately when a requirement is missing or architecture is blocked.
Known limitations
Does not perform implementation itself. Requires sibling skills to be present in the workforce directory to function. Cannot manage external SaaS trackers like Jira.
How to use
Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scanned
Verified clean 1 month ago
- Passed all security checks, Safe to install