Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Chargeback Dispute Response · mohitagw15856 | DecimalAI

financeby mohitagw15856

Chargeback Dispute Response

mohitagw15856/chargeback-dispute-response

Source mohitagw15856/pm-claude-skills ↗License MIT

Win a chargeback dispute — read the reason code, assemble the evidence packet, and write the rebuttal that actually persuades the bank.

SkillScore78Maintained

Automatic

Finance & Accounting pack

Last verified 8/31/2026

What it changes

Passed 18 of 23 test cases with the skill, and 16 of 23 without it.

Without skill70%

With skill78%

Checks it passed that the model alone missed

Rates winnability as Concede.
Includes IP address matching in the evidence table.
Includes top section header starting with '### Chargeback Response —'.

3 fixed15 passed either way4 still failing1 broke

Pass-rate lift: +9 pp
Tokens: +31%
Agent turns: 0%
Response time: -3%

Model gemini-3.6-flashMeasured 8/31/2026v118 of 23 cases passed with skill

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/mohitagw15856-chargeback-dispute-response/SKILL.md https://app.decimal.ai/s/mohitagw15856-chargeback-dispute-response/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add mohitagw15856-chargeback-dispute-response, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/mohitagw15856-chargeback-dispute-response/ and pins it in .decimal/skills.lock

decimalai skills add mohitagw15856-chargeback-dispute-response

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/mohitagw15856-chargeback-dispute-response/SKILL.md https://app.decimal.ai/s/mohitagw15856-chargeback-dispute-response/SKILL.md

Raw file: /s/mohitagw15856-chargeback-dispute-response/SKILL.md ↗ · machine-readable summary at /s/mohitagw15856-chargeback-dispute-response.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 3

Prompt

We have a dispute for $600 with Visa reason code 10.4 (Other Fraud - Card-Absent Environment) dated Sep 20 for an online course. At checkout, AVS street/zip matched, CVV passed, and IP address matched the billing state. Customer access logs show 18 hours of video watched and 75% module completion over three weeks. We have no prior chargebacks from this user. Please decode the reason code into simple terms and evaluate dispute winnability. Output a formatted checklist table mapping necessary evidence to whether we have it and how it is labeled. Draft the representment letter including transaction facts, evidence arguments referencing attached documents, terms accepted at purchase, and a formal closing asking to overturn the chargeback. Conclude with advice on stopping this friendly-fraud pattern.

What was checked, without and with the skill
Checked	Without skill	With skill
Evaluates dispute winnability as Strong.	Passed	Passed
Cross-references course consumption logs as evidence of cardholder participation.	Passed	Passed
Includes top section header starting with '### Chargeback Response —'.	Missed	Passed

case-03 · gemini-3.6-flash · 8/31/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md67 linessingle file

Chargeback Dispute Response

Chargebacks are won or lost on matching the specific reason code to the specific evidence the card networks accept — not on how unfair it feels. This decodes the code, tells you exactly which proof wins that category (delivery confirmation, an AVS match, the signed terms, prior usage), assembles the rebuttal in the structure banks expect, and — importantly — tells you when a dispute is unwinnable so you don't waste the representment.

Card-network rules and win rates vary by processor and code. This structures the strongest representment; it can't guarantee the issuer's decision.

What This Skill Produces

The reason-code decode — what the customer/bank is actually claiming, in plain terms
The evidence checklist — the specific proof that wins this code (not a generic pile)
The rebuttal letter — a structured, factual representment matched to the claim
The winnability read — fight, or concede and save the fee — an honest call

Required Inputs

Ask for these if not provided:

The reason code — the code and category (fraud, product not received, not as described, subscription/cancelled, duplicate)
The transaction — amount, date, product/service, and your records
Your evidence — what you can actually document: delivery/tracking, AVS/CVV match, IP/device, communications, terms accepted, usage logs, refund policy
History — prior chargebacks from this customer, and your processor

Framework: Match Evidence to the Code

The code dictates the fight. "Product not received" is won with delivery proof; "fraud" with AVS/CVV/IP + usage; "not as described" with the listing + comms; "subscription cancelled" with the cancellation policy and usage after the claimed date.
Compelling evidence, not a story. Card networks weigh specific artifacts; emotion and unfairness arguments lose.
Facts, dated, and cross-referenced. Every claim in the rebuttal points to an attached artifact.
Concede the unwinnable. True fraud with no AVS match, or no delivery proof, is a lost representment — take the loss, save the time, and address prevention.
Prevent the next one. Note the pattern (friendly fraud, unclear billing descriptor, no delivery capture) so it stops recurring.

Output Format

Chargeback Response — [reason code] · [$amount]

Claim in plain terms: … · Winnability: Strong / Worth trying / Concede — why

Evidence checklist (for this code)

Required proof	Have it?	Attached as

Rebuttal letter

Opening: transaction facts (date, amount, what was sold)

Track record

Benchmark

18 / 23

test cases passed, vs 16 / 23 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

mohitagw15856 / pm-claude-skills

skills/chargeback-dispute-response/SKILL.md

View on GitHub ↗

License MITImported 8/10/2026Version v1

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Chargeback Dispute Response

What This Skill Produces

Required Inputs

Framework: Match Evidence to the Code

Output Format

Chargeback Response — [reason code] · [$amount]

Evidence checklist (for this code)

Rebuttal letter

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills