break-trace — Claude Code skill: trace reconciliation breaks
The skill traces a reconciliation break to its source entries and states what differs. Use it after gl-recon has classified the break.
npx skills add anthropics/financial-services --skill break-trace -a claude-codeWhat break-trace does
break-trace takes a single break row and checks both sides of the reconciliation. It uses the internal-gl MCP to retrieve the journal entry or posting, then the subledger MCP to fetch the matching transaction. It compares posting dates, FX rates, account mapping, and the signs of quantities and amounts.
The result is a root-cause statement and JSON with the break key, owner, expected clear date, and action. Actions include adjusting a break, but the skill does not post adjustments; only the resolver does that. That boundary matters if you expect the agent to change ledger entries on its own.
Run it after gl-recon classifies a break. The agent needs access to internal-gl MCP and subledger MCP to retrieve the records for tracing.
What it helps with
- trace a break row to its originating GL posting
- match a ledger entry to its subledger transaction
- identify a date, FX rate, or account-mapping difference
- return a root-cause statement and next action as JSON
Example prompts for the agent
- Trace this break to its source entries and state the root cause.
- Compare the GL posting with the matching subledger transaction and return the result as JSON.
- Check whether this break comes from the posting date, FX rate, or account mapping.
How to install break-trace
1Skills CLI (Vercel)
npx skills add anthropics/financial-services --skill break-trace -a claude-code2Manually: put the files in place
~/.claude/skills/break-trace/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/break-trace/
1Skills CLI (Vercel)
npx skills add anthropics/financial-services --skill break-trace -a codex2Manually: put the files in place
.agents/skills/break-trace/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add anthropics/financial-services --skill break-trace -a cursor2Manually: put the files in place
.cursor/skills/break-trace/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install anthropics/financial-services break-trace2Manually: put the files in place
.github/skills/break-trace/SKILL.md1Skills CLI (Vercel)
npx skills add anthropics/financial-services --skill break-trace -a gemini-cli2Platform CLI
gemini skills install https://github.com/anthropics/financial-services --path plugins/agent-plugins/gl-reconciler/skills/break-traceManually: put the files in place
.devin/skills/break-trace/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add anthropics/financial-services --skill break-trace -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set51
The whole set →Similar
Questions about break-trace
How do I install break-trace in Codex?
Run npx skills add anthropics/financial-services --skill break-trace -a codex. The skill file is at .agents/skills/break-trace/SKILL.md.
What does break-trace do?
It follows a reconciliation break to the originating GL posting and matching subledger transaction, then states the cause and a proposed action.
Can break-trace post an adjustment?
No. It diagnoses the break; only the resolver writes adjustments.
Source text
SKILL.md · Apache-2.0
---
name: break-trace
description: Root-cause a reconciliation break to its source transaction or posting — follow the audit trail from the break row back to the originating entry on each side and state what differs and why. Use after gl-recon has classified a break.
---
# Root-cause a break
Given a single break row (key, GL values, subledger values, bucket, likely cause), trace it to source and produce a root-cause statement.
## Trace path
1. **Pull the GL side** — via the internal-gl MCP, fetch the journal entry or posting that produced this GL line: entry id, posting date, source system, batch id, preparer.
2. **Pull the subledger side** — via the subledger MCP, fetch the matching transaction: trade id, trade/settle dates, counterparty, source feed, FX rate used.
3. **Diff the attributes** — line up posting date, FX rate/date, account mapping, quantity sign, amount sign. The differing attribute is usually the cause.
## Cause → statement
Write the root cause as a single sentence in the form **"⟨side⟩ ⟨did what⟩ because ⟨reason⟩"**, e.g.:
- "GL posted on settle date (T+2) while subledger posted on trade date — timing break, will clear on 2026-05-07."
- "Subledger used WM/R 4pm rate; GL used Bloomberg close — FX break of 12 bps on the base amount."
- "Security ABC123 maps to GL account 11420 in the mapping table but the subledger fed 11410 — mapping break, raise to reference-data."
- "Subledger posted the trade twice (trade ids 88412 and 88419 are duplicates) — duplicate post, suppress 88419."
## Output
For each traced break, return:
```json
{
"key": "...",
"root_cause": "one sentence as above",
"owner": "ops | reference-data | accounting | upstream-system",
"expected_clear_date": "YYYY-MM-DD or null",
"action": "monitor | adjust | raise-ticket | suppress"
}
```
Only the resolver writes adjustments — this skill diagnoses, it does not post.
© anthropics · Apache-2.0 · anthropics/financial-services