Dispute Respond logo

Dispute Respond

Organization
runxhq
dispute-respond

Prepare and explicitly file an evidence-bound payment dispute response using native receipt proof, approval, provider execution, and independent readback.

Overview

Publisherrunxhq
Repositoryrunx
Skill namedispute-respond
Stars
105
Forks
109
Bundled files
6
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 6 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by runxhq on GitHub. Read the source before you install it.

Installation

Install the Dispute Respond AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/runxhq/runx/tree/main/skills/dispute-respond
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/runxhq/runx.git /tmp/runx
mkdir -p .claude/skills
cp -r /tmp/runx/skills/dispute-respond .claude/skills/dispute-respond
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Dispute Respond in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Dispute Respond on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Dispute Respond is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Dispute Respond

Turn one provider dispute event and the exact Runx payment history behind it into a reviewable response posture. A good dispute packet does not merely tell a persuasive story: it proves which charge and refund receipts were consulted, which external evidence may be cited, what remains unknown, and why the operator should accept, contest, review, or acknowledge an existing refund.

The default file-response runner prepares the evidence-bound response, files it through the selected configured provider after one exact approval, and requires independent readback. respond remains the explicit draft-only runner, while file consumes an admitted prior response packet without repeating the evidence work. No runner closes a dispute, moves money, or invents provider acceptance.

When to use it

Use dispute-respond after a provider has created a bounded dispute event and the original charge should exist in Runx receipt history. Use refund when the operator chooses a new reversal rather than a dispute response. If the original settlement cannot be proven, gather that evidence before contesting.

How it works

  1. Normalize provider, dispute, charge, amount, currency, reason, and optional response deadline.
  2. Resolve the exact original charge and every prior refund by content-addressed Runx receipt id. Caller-supplied receipt rows or summary projections are not accepted as substitutes.
  3. Require complete signature verification, a hosted or CI issuer, and redacted native detail for each linked receipt tree. A receipt signed by a runx home's own local key proves only that operator's word.
  4. Admit external evidence only as bounded locator, SHA-256 digest, kind, and summary. Raw provider payloads and credentials remain outside the packet.
  5. Draft one posture using only admitted receipt ids and evidence locators.
  6. Deterministic finalization rejects invented citations, effect claims, and any completion state the evidence cannot support.
  7. When the operator deliberately invokes file, Runx validates the unchanged ready packet, binds it to a SHA-256 digest, obtains approval, and calls the configured provider under the narrow dispute.file scope.
  8. Runx projects only the filing reference, dispute id, request digest, and status from the provider result, then performs an independent dispute.read and requires those fields to match before reporting success.

A contest posture must cite delivery or consent evidence. A verified prior refund forces refund_already_sent or operator_review; the model cannot ignore it to produce a stronger contest narrative.

Inputs and result

The input includes the normalized dispute event, exact original receipt id, prior refund ids, and bounded external evidence. The result contains the posture, response summary, cited receipt ids and evidence locators, open questions, validation, and an exact provider-filing handoff for later review.

ready_for_review means only that the packet is complete enough for an operator. The preparation result always reports filing.status: not_filed, provider_status: not_called, and approval_status: not_requested.

Use the named file runner only after reviewing that packet. Supply the exact packet, the configured provider name, and one stable idempotency key for that filing. Runx injects the key into the provider request, rejects a conflicting nested key, refuses a packet whose decision or filing state has drifted, and does not accept the mutation response as proof by itself. A successful run must also return matching dispute.read evidence. Provider credentials remain in the configured adapter boundary; they never belong in skill inputs or output.

Stop conditions

  • Return needs_more_evidence for missing, invalid, locally signed, or incompletely verified receipt trees.
  • Refuse caller-authored receipt detail, raw provider evidence, credentials, or citations not present in the admitted set.
  • Do not contest without evidence of delivery, consent, or the provider-specific fact on which the response depends.
  • Do not file a packet that is incomplete, failed validation, already marked as filed, or different from the packet approved by the operator.
  • Do not claim filing from the mutation response alone. Require the independent provider readback, and do not equate a filed response with dispute acceptance, closure, settlement, or a refund.

Example

A dispute alleges an unrecognized charge. Native history proves the charge but contains no consent evidence. The packet may recommend operator_review and name the missing evidence; it should not contest by inference. If a verified refund receipt already covers the amount, the posture becomes refund_already_sent and cites that exact receipt. If the operator later invokes file, the approved packet—not a rewritten summary—is what reaches the provider, and matching provider readback is the completion proof.

Agent task contract

dispute-response-draft

Choose accept, contest, refund_already_sent, needs_more_evidence, or operator_review from the admitted dispute, evidence locators, and native redacted receipt detail. Cite only exact admitted ids and locators. Never claim filing, provider acceptance, closure, settlement, or a refund not proven by the packet.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Dispute Respond AI skill do?

Prepare and explicitly file an evidence-bound payment dispute response using native receipt proof, approval, provider execution, and independent readback.

Why use Dispute Respond on TypingMind?

Because you install it once and use it with any model. Dispute Respond is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Dispute Respond in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/runxhq/runx/tree/main/skills/dispute-respond. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Dispute Respond?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Dispute Respond?

As many as you like. As long as a model supports skills, you can use Dispute Respond with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Dispute Respond AI skill free?

Yes. It is published on GitHub by runxhq under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇