OrgX Skills

Agent skills for OrgX - the AI-powered operations platform.

Installation

# Install all OrgX skills
npx add-skill useorgx/skills

# Install a specific skill
npx add-skill useorgx/skills/morning-briefing

# List available skills
npx add-skill useorgx/skills --list

Available Skills

Skill	Description
morning-briefing	Daily OrgX briefing with pending decisions, blocked tasks, and agent status
initiative-kickoff	Create complete initiatives from a one-line goal
bulk-create	Create multiple tasks from markdown checklists
nightly-recap	End-of-day activity summary
orgx-capability-mindset	Shared Software 3.0 judgment layer: verifier gates, agent-native outputs, jagged-risk routing, adversarial review, and MCP progress ledger
orgx-quality-bar	How OrgX judges artifacts: four-lens verification, domain layer stacks, the AQ 0.85 gate, artifact type codes and the mistype guard, GTM sendability checks, and the rework loop
orgx-design	Canonical OrgX product design system: full-surface audits, app and widget patterns, responsive/accessibility verification, cut protocol, and cross-tool drift control
product-agent	100x product domain expert: PRDs, initiatives, canvases, research briefs, competitive analyses, prioritization matrices, pivot evaluations, dashboard specs, launch readiness
engineering-agent	100x engineering domain expert: RFCs, ADRs, code reviews, postmortems, tech debt inventories, capacity plans, runbooks, migration playbooks, dependency audits, perf budgets
marketing-agent	100x marketing domain expert: campaign briefs, content packs, nurture sequences, positioning docs, messaging matrices, competitive narratives, launch plans, analyst briefs, community strategies
sales-agent	100x sales domain expert: battlecards, MEDDIC scorecards, outreach sequences, territory plans, QBR decks, deal reviews, win/loss analyses, pricing proposals, partner pitches
design-agent	100x design domain expert: WCAG audits, design tokens, component docs, interaction specs, UX research plans, critiques, motion specs, dark mode audits, breakpoint maps
operations-agent	100x operations domain expert: incident analyses, playbooks, budget controls, capacity plans, vendor evaluations, SLO proposals, chaos test plans, migration checklists, on-call audits
orchestrator-agent	100x orchestration expert: initiative plans, delegations, synthesis reports, retrospectives, dependency audits, resource allocations, risk registers, stakeholder updates, program status
initiative-protocol	Initiative lifecycle creation, launch, monitoring, completion
milestone-protocol	Milestone tracking, risk flags, completion
workstream-protocol	Workstream execution lifecycle and blockers
task-protocol	Task lifecycle execution and status reporting

Requirements

These skills require the OrgX MCP server to be configured:

{
  "mcpServers": {
    "orgx": {
      "type": "http",
      "url": "https://mcp.useorgx.com/mcp"
    }
  }
}

Founder/Team Artifact Contract

All role agents inherit the shared orgx-capability-mindset contract. Agents should first identify whether the workspace is acting like an early founder, founder-led company, or operating team, then produce the practical artifact that moves the company forward now.

When attaching artifacts or receipts through OrgX MCP, use the canonical cross-domain artifact type where possible:

Agent	MCP artifact_type examples
Orchestrator	orchestration.next_initiative
Engineering	eng.pull_request, eng.deploy_proof, eng.structured_blocker
Sales	sales.strategy, sales.icp_offer_sequence, sales.follow_up_sequence, sales.send_plan
Marketing	marketing.launch_asset, marketing.channel_hypothesis, marketing.offer_package
Product	product.customer_discovery, product.prd, product.pricing_hypothesis, product.decision_record
Design	design.audit, design.component_spec, design.token_package
Operations	ops.operator_brief, ops.runbook, ops.budget_envelope, ops.incident_status

Attach agent_type, company_stage, business_outcome, owner, review_date, and verification metadata so OrgX can connect artifacts to proof, recommendations, morning briefs, and skill improvement. Use local validator --type names inside each skill; use the MCP artifact_type values when saving the artifact to OrgX.

Quality Bar

OrgX verifies every attached artifact with a four-lens system (judged, measured, observed, outcome) and gates shipping at AQ 0.85. The artifact_type you declare selects which domain layer stack judges the work, and below-gate artifacts come back as rework with layer-specific feedback. orgx-quality-bar documents the full system — every domain agent's Quality Bar section is derived from it.

Verification

Run the repo-level drift check before opening a PR or shipping skill changes:

python3 scripts/check_skill_tool_drift.py

The check fails on deprecated OrgX tool names, unknown mcp__orgx__* references, and high-value workflow gaps such as spawning without orgx_spawn or opening plan sessions without orgx_plan.

Skill Format

Each skill follows Anthropic's skill format:

skill-name/
├── SKILL.md      # Required: Instructions with YAML frontmatter
├── skill.toml    # Optional: OrgX registry metadata
└── references/   # Optional: Additional documentation

Contributing

1. Fork this repository
2. Create a new skill directory with a SKILL.md file
3. Follow the Anthropic skill authoring best practices
4. Submit a pull request

License

MIT