agentleFS
DevinoSolutionsDevinoSolutions/claude-plugins/skills/cleaning-up-bookkeeping/SKILL.md

cleaning-up-bookkeeping

Skill0 starsChanged 9 days agoMITView on GitHub

Tidies up SuperBooks transactions and receipts: categorizes or recategorizes transactions, fixes a transaction's category or status, finds uncategorized spending, matches receipts and bills in the inbox to transactions, and finds documents. Use when the user asks to 'categorize last month's transactions', 'match the receipts in my inbox', 'find uncategorized expenses', 'fix this transaction's category', 'reconcile my receipts', or 'find the receipt for this purchase'.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Operations listed by name: call them directly, for example superbooks:invoices_list.

    line 16
  2. Very specific:

    If the connection lists the operations by name instead (the per-tool surface), call them directly without the external_ prefix.

    line 24
  3. Very specific:

    If a category the user wants does not exist, offer to create it with external_categories_create first.

    line 45
  4. Very specific:

    Check the real argument names in the superbooks:search_tools declaration before running this.

    line 61
  5. Distinctive:

    No change without the user's approval of that specific change.

    line 65
  6. Very specific:

    Never guess a category slug.

    line 66

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md
<!-- Generated from plugins/superbooks/skills/cleaning-up-bookkeeping/SKILL.md by scripts/sync-skills.mjs. Edit the plugin copy, then run the script. -->

Bookkeeping cleanup in SuperBooks

Work through transactions that need a category or status, and connect receipts to the transactions they belong to. Every change is shown to the user before it is made.

Connecting and calling the tools

Every operation in this skill comes from the superbooks MCP server, the SuperBooks connector. The server defaults to Code Mode, so the operations below are written as the external_* functions you call inside superbooks:execute_typescript. On a server set to the per-tool surface, the same operation is listed as superbooks:<name> without the external_ prefix, for example superbooks:invoices_list.

  • Only superbooks:search_tools and superbooks:execute_typescript listed (Code Mode, the default): call superbooks:search_tools first, then call each tool inside superbooks:execute_typescript as external_<tool> (for example external_invoices_list); the program must return its result. Scopes, confirmations and gates are identical on both surfaces.
  • Operations listed by name: call them directly, for example superbooks:invoices_list.
  • "Tool not found": switch to the Code Mode route instead of retrying the direct call. If superbooks:search_tools does not return the tool either, its OAuth scope was not granted; the user reconnects SuperBooks and approves that scope.
  • No superbooks tools at all: SuperBooks is not connected yet. Follow the setup-superbooks skill.

Tools you will use

The superbooks MCP server has two tools: superbooks:search_tools (call it first, for example with transaction, categor, inbox, to get the exact input shapes) and superbooks:execute_typescript (run a short program calling the declared external_* functions; the program must return its result).

If the connection lists the operations by name instead (the per-tool surface), call them directly without the external_ prefix.

Operations this skill uses:

  • external_transactions_list: transactions for a period (for example from: "2026-09-01"), with amount, merchant and category_slug.
  • external_transactions_get: one transaction in full.
  • external_categories_list: the team's categories and their slugs. external_categories_create adds one the user asks for.
  • external_transactions_update_category: set a transaction's category by slug.
  • external_transactions_update_status: set a transaction's workflow status.
  • external_tags_list and external_tags_create: labels the user may want.
  • external_inbox_list: incoming receipts and bills waiting to be matched.
  • external_inbox_match: attach an inbox item to a transaction.
  • external_documents_search, external_documents_list, external_documents_get: receipts and documents already stored.

Workflow

  1. Agree the period (default: last calendar month). In one program, fetch external_transactions_list for it and external_categories_list with await Promise.all.
  2. List the transactions that are uncategorized or that the user flagged. For each, propose a category slug from the category list, based on the merchant and amount.
  3. Show the proposals as a table (date, merchant, amount, current category, proposed category) and let the user approve all, some, or edit them.
  4. Apply the approved ones in one program: one external_transactions_update_category call per transaction, each wrapped so one failure does not hide the others. Report what changed and what failed.
  5. For receipts: fetch external_inbox_list, pair each item with the transaction of the same amount and a close date, show the pairs, and apply external_inbox_match for the ones the user approves.
  6. If a category the user wants does not exist, offer to create it with external_categories_create first.

Code Mode pattern

const approved = [{ id: "txn_1", slug: "design_software" }];
const results = await Promise.all(
  approved.map((t) =>
    external_transactions_update_category({ id: t.id, category_slug: t.slug })
      .then(() => ({ id: t.id, ok: true }))
      .catch((e) => ({ id: t.id, error: String(e) })),
  ),
);
return results;

Check the real argument names in the superbooks:search_tools declaration before running this.

Rules

  • No change without the user's approval of that specific change. Batches are fine once the user approves the batch.
  • Never guess a category slug; use one returned by external_categories_list.
  • Deleting transactions, receipts, or categories is not offered here. Destructive operations are off by default for the team; do not ask the user to turn them on.
  • Each credential is limited to 120 calls a minute. For large batches, work in chunks of about 50 and report progress between them.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.