Route an Agoragentic task to the smallest applicable skill. Use when the request involves Agoragentic execution, governance, transaction assurance, proof/receipts, deployment, selling, or integration and the correct branch is not yet known.
Installs into .claude/skills of the current project.
Are you the author of Agoragentic?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/rhein1-agoragentic)
---
name: agoragentic
description: Route an Agoragentic task to the smallest applicable skill. Use when the request involves Agoragentic execution, governance, transaction assurance, proof/receipts, deployment, selling, or integration and the correct branch is not yet known.
---
# Agoragentic Router
Load only the smallest skill that matches the requested job:
- **agoragentic-execute** — preview or execute a bounded capability after explicit approval.
- **agoragentic-govern** — classify side effects, apply policy, and prepare approvals before action.
- **agoragentic-prove** — create or inspect local proof, receipts, evidence references, and reconciliation state.
- **agoragentic-assure** — prepare and evaluate an agent transaction without self-granting authority or moving money.
- **agoragentic-deploy** — prepare Agent OS / Harness deployment-readiness handoffs without provisioning implicitly.
- **agoragentic-sell** — prepare listing/payment readiness without publishing, funding, or activating settlement automatically.
- **agoragentic-integrate** — connect an external host, framework, tool, or specialist engine to Harness governance and receipts.
- **agoragentic-use** — use the Agoragentic marketplace directly from an assistant: discovery, registry, receipts, spend, registration, invocation, federation.
## Routing Rules
1. Prefer one focused skill over loading the full product surface.
2. Check live Agoragentic discovery before claiming provider availability, verification, pricing, payment support, or deployment state.
3. Preview first when a task may spend, publish, deploy, message, mutate trust, or create another side effect.
4. Treat missing policy, identity, cost, approval, or evidence as **BLOCKED** rather than inventing authority or proof.
5. Keep local proof/receipt separate from settlement receipts, certifications, trust endorsements, marketplace verification, and owner approval.
6. Never expose credentials, wallet secrets, raw private prompts, unrestricted private tool output, or private ECF payloads.
## Canonical Discovery
- Skill contract: <https://agoragentic.com/skill.md>
- LLM summary: <https://agoragentic.com/llms.txt>
- API contract: <https://agoragentic.com/openapi.yaml>
- Agent discovery: <https://agoragentic.com/.well-known/agent-card.json>
- MCP card: <https://agoragentic.com/.well-known/mcp/server.json>
- Integration catalog: <https://github.com/rhein1/agoragentic-integrations>
## Advanced Context (Load Only When Needed)
- x402 and payment safety: <https://github.com/rhein1/agoragentic-integrations/tree/main/x402>
- federation and interchange: <https://github.com/rhein1/agoragentic-integrations/tree/main/interchange>
- local memory governance: <https://github.com/rhein1/agoragentic-memory>
- document and workflow contracts: <https://github.com/rhein1/agoragentic-integrations/blob/main/docs/agent-workflow-contracts.md>
- Agoragentic Rust Framework HTTP examples: <https://github.com/rhein1/agoragentic-integrations/tree/main/rust-framework>
- security and disclosure policy: <https://github.com/rhein1/agoragentic-integrations/blob/main/SECURITY.md>
This no-spend router grants no deployment, publication, wallet, trust, ranking, credential, or hosted-memory authority.
