agent-agentic-payments

Agent skill for agentic-payments - invoke with $agent-agentic-payments

2.22x

Quality

32%

Does it follow best practices?

Impact

100%

2.22x

Average score across 3 eval scenarios

Securityby

Risky

Do not use without reviewing

Optimize this skill with Tessl

npx tessl skill review --optimize ./.agents/skills/agent-agentic-payments/SKILL.md

Evaluation results

100%

65%

AI Shopping Bot Payment Authorization Setup

Mandate creation and payment authorization workflow

Criteria

Without context

With context

Uses create_active_mandate

100%

amount_cents in cents

100%

Correct period value

100%

Correct kind value

100%

Merchant restrictions set

25%

100%

Uses sign_mandate

40%

100%

Uses verify_mandate

50%

100%

Uses authorize_payment

50%

100%

authorize_payment amount in cents

100%

Sequential workflow order

100%

Includes metadata in authorize_payment

20%

100%

Without context: $0.2891 · 1m 51s · 16 turns · 22 in / 5,702 out tokens

With context: $0.2633 · 1m 32s · 14 turns · 431 in / 4,102 out tokens

100%

85%

Enterprise Procurement: Multi-Agent Approval Workflow

Multi-agent Byzantine consensus for high-value transactions

Criteria

Without context

With context

Uses request_consensus

100%

required_agents is an array

100%

threshold is numeric

100%

threshold less than total agents

100%

timeout_seconds present

100%

Uses verify_consensus

100%

verify_consensus signatures array

100%

Consensus before authorization completes

50%

100%

Large amount in cents

100%

Byzantine reasoning documented

100%

Without context: $0.8257 · 3m 33s · 30 turns · 792 in / 11,937 out tokens

With context: $0.3947 · 2m 3s · 20 turns · 22 in / 5,772 out tokens

100%

13%

Payment Mandate Audit and Revocation System

Mandate lifecycle management and audit trail

Criteria

Without context

With context

Uses list_mandates

100%

list_mandates status values

70%

100%

Uses get_payment_status

100%

Uses revoke_mandate

100%

Revocation reason present

100%

Audit trail logging

100%

audit_report.md produced

100%

Compliance statement included

100%

Multiple status categories checked

100%

agent_id format correct

100%

Without context: $0.3475 · 1m 46s · 19 turns · 26 in / 5,611 out tokens

With context: $0.4824 · 2m 8s · 22 turns · 281 in / 7,435 out tokens

Repository: ruvnet/claude-flow
Commit: b2618f9

Evaluated: 17 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

AI Shopping Bot Payment Authorization Setup Enterprise Procurement: Multi-Agent Approval Workflow Payment Mandate Audit and Revocation System

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.