Foreword¶

When friends borrow money and don’t return it, colleagues fail to transfer their share for split bills, or clients drag out final payments—asking for money often takes a greater mental toll than owing it. Writing messages yourself often leaves you stuck between “fear of hurting the relationship” and “fear of not being taken seriously.” Meanwhile, asking a large language model vaguely often yields single-paragraph replies that are either too soft or too harsh, lacking a escalating rhythm based on relationship, amount, and frequency.

Below is an introduction to the SkillHub community Skill “Payment Reminder Scripts” (@user_70c2f807/payment-reminder-scripts). It organizes debt collection into reusable execution steps and a fixed output structure, allowing the Agent to first clarify the situation within the conversation, then provide multiple sets of scripts, rebuttal handling, and a follow-up schedule. The Skill is marked as requires_api_key=false and does not depend on external API keys.

What is this¶

“Payment Reminder Scripts” is an Agent Skill targeting business-ops scenarios. It adopts the universal SKILL.md format, with the current version at 1.0.0, sourced from the community, and published by user_70c2f807.

When users mention debt collection, asking for payment, money owed, or intents like payment reminder, collect debt, or money owed in English contexts, the Skill guides the assistant to generate communication plans that are tactful yet clear in stance. It does not cover legal debt litigation, debt arbitration, lending intermediation, or non-compliant collection practices.

SkillHub Page: Payment Reminder Scripts. The resource page shows a rating of 100000 and 895 downloads (subject to SkillHub display).

Core Features¶

The following content is derived from the Skill body and represents the capability boundaries the Agent will follow after installation.

Scenarios and Strategies¶

1. Adjust tone and strategy based on relationship type (friends, colleagues, clients, relatives, roommates, partners, etc.).
2. Match collection intensity based on amount tiers (small split bills, medium loans, large cooperation payments).
3. Provide four-level progressive scripts: from hinting, reminding, urging, to “ultimatum.”
4. Design natural entry scenarios to reduce the perception of being “petty.”
5. Provide response patterns for common excuses (tight on cash, forgot, in a few days, read but not replied, etc.).
6. Suggest collection intervals and bottom lines (when to consider legal channels, etc.).

Execution Flow (Four Steps)¶

The Skill requires the assistant to complete the following in order:

1. Understand the collection situation: Relationship, amount and duration, previous collection attempts and reactions, user’s bottom line (full amount / installments / flexible for small amounts).
2. Formulate collection strategy: Relationship weight, communication method (hint / opportunistic / direct / clear stance), channel and timing, anticipated excuses.
3. Generate script plans: Typically 2–3 complete sets of scripts ranging from mild to firm, including introduction, core expression, excuse handling, and a follow-up schedule.
4. Preventive suggestions: Risk control tips such as transfer notes, agreed repayment dates, IOUs, or record retention.

Output Structure¶

Replies are organized according to the Markdown template agreed upon in the Skill, generally including: Debt relationship and amount, duration and collection level, situation analysis, multiple script sets, response comparison table, schedule, and precautions/preventive suggestions. Refer to the “Output Format” section in SKILL.md for specific fields.

Explicit Exclusions¶

The Skill states limitations in the body text, which the assistant should explain to the user in advance:

• Does not provide legal debt collection services; large amounts (the body text example is over 50,000 or involving contract disputes) should consult a lawyer or pursue legal channels.
• Does not guarantee collection results; scripts are communication tools only.
• Refuses threatening, harassing, publicly shaming, or privacy-infringing collection methods.
• Does not store user lending and relationship details (conversational level use only, no persistence promise).

Installation and Activation¶

The installation method provided on the official page is: Let the AI refer to the SkillHub installation documentation to install the coordinate @user_70c2f807/payment-reminder-scripts. Documentation address: https://skillhub.cn/install/skillhub.md.

1. Check or Install SkillHub CLI¶

command -v skillhub && skillhub --version

If not installed, you can execute according to the documentation (full installation includes CLI and default Skills):

curl -fsSL https://skillhub-1388575217.cos.ap-guangzhou.myqcloud.com/install/install.sh | bash

To install only the CLI:

curl -fsSL https://skillhub-1388575217.cos.ap-guangzhou.myqcloud.com/install/install.sh | bash -s -- --cli-only

2. Install to the current Agent’s skills directory¶

During installation, you must use --dir to point to the skills directory recognized by the current client; otherwise, it defaults to ./skills/ and may not be loaded.

skillhub install @user_70c2f807/payment-reminder-scripts --dir <your skills directory>

Common paths (from official installation documentation):

• Cursor: ~/.cursor/skills/
• Claude Code: ~/.claude/skills/
• Codex: ~/.codex/skills/ or .agents/skills/ under the project

After installation, restart or refresh the Skill index as required by your client. This article has not verified configuration differences for Codex CLI, Claude Code, and Cursor post-2026; the above is based solely on the directories listed in the SkillHub documentation.

Typical Usage¶

The Skill provides a reproducible conversation form in the “Quick Start” section. Once installation is complete, it can be triggered directly using natural language.

Example 1: Long-overdue friend loan

User: A friend borrowed 3000 yuan from me two months ago and hasn’t returned it. How should I bring it up?

Expected Assistant Behavior: Clarify relationship closeness and whether it has been asked for before; provide scripts like “Natural Introduction” or “Direct & Friendly,” accompanied by excuse handling and a follow-up rhythm.

Example 2: Small AA payment from colleague

User: A colleague said we’d split the bill after dinner last time but didn’t transfer the money. It’s only 200 yuan and I feel awkward asking.

Expected Assistant Behavior: Provide low-pressure small-amount scripts, such as combining it with a reminder about the dinner context, payment tools, or bringing it up naturally during the next dinner.

Example 3: Client’s final payment

User: The client’s project final payment has been delayed for two months. I asked once and they said it’s going through the process.

Expected Assistant Behavior: Provide business-oriented scripts and follow-up strategies, potentially including email templates, WeChat follow-ups, and escalation nodes (e.g., when to consider a collection letter), distinguishing between formal processes and daily private chat scenarios.

Example 4: Multiple delays

User: I’ve asked three times and they keep saying “in a few days.” What should I do?

Expected Assistant Behavior: Differentiate between “forgetfulness” and “intentional delay,” suggest pinning vague promises to specific dates, and clarify the bottom line on stance when invalid.

Trigger Keywords (from Skill header): Debt collection, asking for payment, money owed, payment reminder, collect debt, money owed, etc.

Applicable Scenarios & Notes¶

Who is it for¶

• Individuals or small teams needing written scripts and schedules rather than one-off improvised replies.
• Handling common interpersonal debts (friends/colleagues/clients/relatives), mainly for small to medium amount communication as described in the Skill.
• Developers or operators already using SkillHub installation flows in clients supporting Skill directories (Cursor, Claude Code, Codex).

Limitations and Misuse¶

The “Inapplicable Scenarios” in the Skill also reminds: Do not publicly hint at collection in group chats or朋友圈 (Moments); do not accuse emotionally; avoid starting with the hardest stance immediately, nor should one offer endless soft collections without escalation.

When information is incomplete, the assistant should ask follow-up questions according to the “Error Handling” table: Debtor relationship, amount, duration of delay, previous collection attempts, etc.

Conclusion¶

“Payment Reminder Scripts” converges debt collection from awkward improvised chatting into verifiable situational Q&A, tiered scripts, and follow-up rhythms. It is suitable for hanging in daily Agent workflows to handle high-frequency, small-stakes communication difficulties. It does not replace lawyers or financial processes, but offers a workable middle layer between saving face and maintaining a clear stance.

For installation and details, see the SkillHub official page: https://www.skillhub.cn/skills/user_70c2f807/payment-reminder-scripts. Coordinate: @user_70c2f807/payment-reminder-scripts, Version 1.0.0.