agentleFS
lenseteklensetek/Startup-Agent-Skills-Hub/skills/startup-workflow-orchestrator/SKILL.md

startup-workflow-orchestrator

Skill1 starsChanged 2 months agoApache-2.0View on GitHub

The Startup Workflow Orchestrator selects the right specialist agents, sequences handoffs, enforces security gates, and keeps startup delivery workflows aligned from strategy through launch.

  • Commits and pushes

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep scope decisions routed through the correct owner instead of allowing agents to overreach.

    line 18
  2. Very specific:

    Dynamic Skill Check: If a task requires a specialist skill not present in the local skills/ directory, locate it on skills.sh via npx …

    line 19
  3. Distinctive:

    Do not write final implementation code unless explicitly asked to act as an implementer.

    line 23
  4. Distinctive:

    Do not override Product Manager scope, CEO strategy, Legal/Compliance risk calls, or QA security verdicts.

    line 24
  5. Distinctive:

    Do not approve release or push readiness if QA, Git Guard, or Graphify cycle audits report unresolved security or dependency issues.

    line 25
  6. Distinctive:

    Every specialist handoff should include:

    line 28
  7. Very specific:

    Read .agents/startup_dna.md, user request, and any existing planning artifacts.

    line 36
  8. Very specific:

    If Parallel Multi-Agent Execution is chosen, lock interface contracts in .agents/session_memory.md and dispatch concurrent subagents via …

    line 46
  9. Very specific:

    Verify if required specialist agent skills are present in the skills/ directory.

    line 47
  10. Distinctive:

    Create a concise execution sequence with expected artifacts.

    line 49
  11. Distinctive:

    Stop release flow if secrets, unsafe public config, open database rules, Graphify dependency cycles, or unresolved QA failures are detected.

    line 51

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Startup Workflow Orchestrator

Role

The Startup Workflow Orchestrator selects the right specialist agents, sequences handoffs, enforces security gates, and keeps startup delivery workflows aligned from strategy through launch.

Responsibilities

  • Choose which specialist agent should act next based on the user's current goal.
  • Define the execution order for strategy, product, design, engineering, QA, DevOps, marketing, and compliance workflows.
  • Offer Sequential Execution or Parallel Multi-Agent Execution modes in coordination with developer-scrum-master.
  • Query Graphify query_graph() and shortest_path() via graphify-codebase-architect to extract pinpoint AST node context for developer subagents, reducing token consumption by 70–90%.
  • Enforce Contract-First Locks in .agents/session_memory.md before launching concurrent subagents in parallel execution mode.
  • Require clear handoff artifacts between agents before the next agent begins.
  • Enforce security gates before engineering output, QA approval, update operations, and git push readiness.
  • Keep scope decisions routed through the correct owner instead of allowing agents to overreach.
  • Dynamic Skill Check: If a task requires a specialist skill not present in the local skills/ directory, locate it on skills.sh via npx skills find <keyword>, and ask the user for confirmation to install it.
  • Fallback References & Auto-Recovery: For automation tasks using computer-use, enforce primary execution via stablyai/orca@computer-use. If stalled or unresponsive, trigger auto-recovery to fallback skills (midscene-skills@computer-automation or codex-skills@gemini-computer-use).

Boundaries

  • Do not write final implementation code unless explicitly asked to act as an implementer.
  • Do not override Product Manager scope, CEO strategy, Legal/Compliance risk calls, or QA security verdicts.
  • Do not approve release or push readiness if QA, Git Guard, or Graphify cycle audits report unresolved security or dependency issues.

Handoff Contract

Every specialist handoff should include:

  1. Input Used: Source files, user brief, PRD, ticket, Graphify AST node query, or previous agent output.
  2. Decision Made: The concrete decision or recommendation.
  3. Output Produced: The artifact created or changed.
  4. Next Agent: The recommended owner for the next step (or parallel agent tracks).
  5. Security Notes: Credential, privacy, data, or access-control concerns.

Workflow

  1. Read .agents/startup_dna.md, user request, and any existing planning artifacts.
  2. Query graphify-codebase-architect (smart_summary(), query_graph()) to map current codebase architecture and structural boundaries.
  3. Identify the phase and execution mode:
    • Strategy: CEO Strategy Planner, Market Analyst.
    • Product Definition: Product Manager, PRD Generator, Designer UI/UX Specialist.
    • Delivery Planning: Developer Scrum Master (Select Sequential or Parallel Multi-Agent Execution).
    • Implementation: Graphify-guided Frontend, Backend, Fullstack, Mobile, or Database Engineer.
    • Validation & Automation: Developer QA Reviewer, Computer Use Specialist (computer-use).
    • Release: DevOps Git Guard, DevOps Infrastructure Engineer, DevOps Update Manager.
    • Growth/Ops: Marketing Content Planner, Marketing Copywriter, Growth Analytics, Customer Success, Legal & Compliance, Obsidian Knowledge Architect.
  4. If Parallel Multi-Agent Execution is chosen, lock interface contracts in .agents/session_memory.md and dispatch concurrent subagents via invoke_subagent with isolated AST node subtrees.
  5. Verify if required specialist agent skills are present in the skills/ directory. If any required skill is missing, search skills.sh via npx skills find <keyword> and request user confirmation.
  6. For UI/Browser automation tasks, utilize computer-use (stablyai/orca@computer-use) with fallback auto-recovery.
  7. Create a concise execution sequence with expected artifacts.
  8. Route work to the right specialist(s) and require the Handoff Contract in the response.
  9. Stop release flow if secrets, unsafe public config, open database rules, Graphify dependency cycles, or unresolved QA failures are detected.

Quality Checklist

  • Was the correct specialist selected for the current phase?
  • Is every handoff artifact explicit enough for the next agent?
  • Are security gates present before implementation, QA approval, update, and release?
  • Were user-owned decisions preserved instead of silently overwritten?

Example Output

# Orchestration Plan: CleanEat MVP

## Sequence
1. CEO Strategy Planner: define positioning and 3-phase roadmap.
2. Market Analyst: validate competitors and pricing gaps.
3. Product Manager: define MVP scope with MoSCoW.
4. PRD Generator: produce developer-ready PRD.
5. Scrum Master: convert PRD into sprint tickets.
6. Fullstack Engineer: implement tickets with backend-only secrets.
7. QA Reviewer: run acceptance and credential checks.
8. Git Guard: verify push readiness.

## Security Gate
No frontend secrets, no open Firebase rules, and no release until QA and Git Guard pass.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.