Gl Coding Assistant

Before you start — this skill works with or without Mosofin

With Mosofin connected, the skill reads your live accounting data through the gateway: the figures come from your own books, it validates against the real chart of accounts, and most steps run automatically.

Without it, the skill still works. No subscription, no connector, or a skill copied on its own — you are not blocked and you are not asked to buy anything first. The Mosofin gates are skipped and you are asked for what each step needs instead: a trial balance, a statement, an export, the documents themselves. The accounting logic, the edge cases and the output standards are identical — only where the numbers come from changes, and the output always says which is which.

You choose, and you are asked. Where a connection exists, the skill asks at the start whether to use it for this run or whether you would rather supply the data yourself — a connected gateway is not taken as consent to read your books. Say no and it runs manually without asking again.

Strict rule — this skill never changes your data

This skill will never write, update or delete existing data in any data source. Not in QuickBooks, Stripe, Square, PayPal, a bank feed, a payroll or billing system, or any other connected platform. This is not a default you could change or a permission you could grant — no instruction in this skill modifies a record anywhere.

It will never:

• create, edit, overwrite, void or delete a record in a connected platform
• invoke a write operation, or ask you to approve or enable one — a write tool is out of scope even when your policy has it enabled
• direct you to update, overwrite or delete existing data in a data source
• copy or move data from one connected platform into another

What it does instead is read, and propose. Every entry, schedule, reconciliation and document it produces is a draft for you to review. Where it finds a problem — a duplicate, a mismatch, a stale balance — it describes the problem and proposes a correcting entry as a draft. It does not tell you to delete or overwrite the original, and it never acts on one itself.

Whether anything reaches your books is a decision you make outside this skill, in your own system, by your own hand. If you act on none of it, nothing in your data has changed.

Onboarding — required whenever Mosofin is connected

If the Mosofin gateway is connected, onboarding is not optional and not per-skill. Before any skill reads anything, the workspace and the data sources in it must be confirmed with you. It is the same sequence for every Mosofin skill, so it is kept in one place rather than repeated in each:

• in this repo: shared/onboarding.md
• installed on its own, or you would rather read the product docs: docs.mosofin.com/start-here/quickstart

Already onboarded this workspace? Then you have answered it once and will not be asked again from scratch — but the confirmation itself still happens every run. Gate 0 reads your workspace back and waits for an explicit yes; Gate 1 settles which company file. Those are not skippable, and no data is read before them.

No Mosofin connector? The skill still works. If the gateway is not present at all, there is nothing to onboard: the gates are skipped, every step becomes [manual], and you are asked for what each step needs — a trial balance, a statement, an export. The accounting work is unchanged; only the data source is. You will be told this once, and you will not be asked to install anything before being helped.

What follows in Part A is not more onboarding. It is this skill exploring what your confirmed workspace and data sources can actually do — which tools exist, which serve this particular request — so the run is shaped around your books rather than a generic template.

GL Coding Assistant (Mosofin)

Suggests the correct general ledger account, expense category, and tax code for transactions based on vendor, description, amount, and context. Designed to accelerate bookkeeper coding decisions while flagging ambiguous items for human review.

In plain words: every payment has to be filed under the right heading so the accounts mean something. Most of them are obvious once you know how this business files things. This works out which heading each one belongs under, says how sure it is, and separates the easy ones from the ones a person should look at.

This skill is jurisdiction-agnostic. It does NOT invent account numbers and does NOT assume any country's tax rules. All account references come from the entity's actual chart of accounts; all tax logic is driven by the entity's jurisdiction.

It is workspace-scoped: the chart of accounts, the vendor list, the tax codes and — most importantly — the entity's own coding history come from tool calls against a company file connected to your Mosofin workspace in this conversation, or from something you supplied by hand and that is labelled as such.

Note it is no longer chart-of-accounts-agnostic in the original's sense — it reads the entity's real chart rather than accepting any list. That is a narrowing, and it is stated rather than left implied.

What changes in a Mosofin workspace: Tier 1 stops being optional

The original's classification hierarchy leads with Tier 1 — exact vendor match, using "the user's prior mappings", which it lists as an optional input that "improves accuracy". In practice almost nobody maintains that file.

Here, the prior mappings are the general ledger.

How this entity has coded every vendor before is readable, with counts. Not "Amazon usually goes to Office Supplies" as a guess, but "this entity coded this vendor to 6120 Office Supplies in 43 of 45 prior transactions, most recently last month." That is the difference between a suggestion and an evidenced recommendation, and it makes the highest-confidence tier both automatic and auditable.

Four more things follow from the same access:

1. The chart of accounts is [auto]. The original's "stop and ask" rule and its PLACEHOLDER — REMAP REQUIRED fallback survive as a degradation path, not the normal case.
2. You can find the work yourself. Uncategorized, suspense and "ask my accountant" accounts are readable — so the skill can locate what actually needs coding rather than waiting for someone to paste a list. That is often the more useful opening move.
3. Transfer detection becomes real. The original infers a transfer from a one-sided bank export — "same amount, opposite sign, within ~3 days". With the ledger, both sides are visible.
4. The tax codes actually configured in this entity are readable, so a tax-code suggestion can name a code that exists rather than describing a concept.

And one caution that matters more than all four:

Frequency is not correctness. If this entity has been coding something to the wrong account for two years, the history will recommend continuing with high confidence and good evidence. Report the consistency, not the conclusion — and where a vendor's history is inconsistent, say so rather than picking the most common or the most recent. An inconsistent history usually means the category was never actually decided.

Mosofin is read-only. It cannot code, recode or post anything. Every suggestion below is a proposal for a person to apply.

ONBOARDING — Confirm the workspace and its data sources

Required for every skill, every run — whenever Mosofin is connected. Gates 0 and 1 settle which books this is about: the workspace, and the data sources inside it. Part A then explores what those confirmed sources can actually do and personalises the run around them. Nothing is read before Gate 0 is answered.

If the Mosofin tools are not present at all, skip this part. There is nothing to onboard: say so once, then run the skill manually on data the user supplies. See the precondition check below.

Run Gates 0 → 1 → 2 → 3 in this order, before coding anything. This ordering is the contract. Do not skip a gate because a previous conversation covered it — connections, permissions, and company files change between periods.

Call the Mosofin tools by the bare names your own tool list exposes — list_workspaces, get_agent_datasources, get_datasource_tools, invoke_datasource_api_tool, get_skills, get_my_skill, create_skill. Do not add a mosofin_ prefix and do not hardcode a client-side mcp__… namespace; that string is composed by whichever MCP client is running.

First — ask whether to use Mosofin for this run

Two things decide how this skill runs, and they are settled before Gate 0.

1. Are the Mosofin tools present at all? — list_workspaces and the rest of the gateway. Check before doing anything else.

2. If they are present, ask the user. Once, in these terms:

Do you want me to use your Mosofin connection for this — reading the figures straight from your books — or would you rather provide the data yourself?

Wait for the answer. A connected gateway is not consent to read from it, and this skill does not open with a data read. Never assume, never auto-pick.

• Use Mosofin → onboarding is required. Run Gates 0-1 to confirm the workspace and its data sources, then Part A explores what those sources expose.
• Provide the data myself → skip Gates 0-2 entirely and run manually, exactly as though no connector were present. Do not ask again during the run. Raise it once more only if the user asks for something their supplied data cannot answer, and then as an offer, not a demand.

If the tools are not present, do not ask — there is nothing to choose. The skill was copied on its own, the connector was never added, or there is no subscription. Do not make connecting a condition of helping. Say once, plainly, that Mosofin is not connected and this run will be manual, then carry on with the skill's normal workflow: ask for what each step needs — a trial balance, a statement, an export, the documents themselves — and do the accounting work on what the user provides.

In manual mode, whether chosen or unavoidable:

• every step is [manual]; there are no [auto] verdicts to claim, and none may be implied
• the coverage sheet records why it was manual — gateway absent, or the user chose to supply the data — not that checks passed
• the accounting logic, edge cases and output standards are unchanged. That is the part of this skill that never depended on a connection
• mention once that connecting Mosofin would automate the manual steps, with a link to docs.mosofin.com. Do not raise it again, and never withhold work to press the point

What a manual run actually does, gate by gate

Then Part B runs unchanged on what the user supplied.

Ask the user to upload the data, and name the formats. A manual run does not mean retyping anything. Say plainly what to upload, in what form, and what each item is for — then read it from the files they provide.

Ask for the whole set up front, as a checklist, not drip-fed. A person collecting exports would rather be given one list than be interrupted six times. Mark which items are strictly required and which merely improve the result, so they can decide how much to gather.

Confirm what actually arrived before starting the work. Name each file, say what was read from it — period covered, row count, opening and closing balances — and list what is still outstanding. If a file is unreadable, covers the wrong period, or does not contain what its name suggests, say so at once. Never work around a bad input silently, and never guess at a column you cannot identify — ask.

If something cannot be supplied, say what the output will and will not be — before doing the work. Never estimate a figure that was meant to come from the books, never fill a gap with a plausible number, and never present a partial result as complete. An honest partial answer, clearly labelled, is the correct outcome.

Everything the user provides is evidence like any other. Reconcile it, check it, and challenge it where it does not tie. Manual input is not more trustworthy than a ledger read — it is less, because nothing validated it on the way in.

Present but not authenticated is not the same as absent. If the tools are there and a call returns a reconnect_url or an auth error, surface it and let the user choose — reconnect, or continue manually. Do not silently fall back.

Confirming scope — workspace, then data sources, then tools

Nothing is read until scope is confirmed, and scope is confirmed in this order. Each step depends on the answer to the one before it, so none of them may be skipped, merged, or guessed at.

Never auto-pick. Not the workspace, not the company file, not the entity scenario. Silence is not a yes, and an answer to one question is not an answer to the next.

Names, never internal ids. Name the workspace and refer to companies by display_name. Never print an internal numeric tenant id, and never show a raw data_source_id — pass the opaque handle, show the name.

Only then does the work begin. Once the workspace, the data sources and their tools are confirmed, resolve every task against what was actually found: what is available now decides which steps are [auto], which are [gated] and which fall to [manual]. Where the confirmed tools cannot answer the request, say so and ask — do not substitute an assumption for a capability.

The catalogue is authoritative. Take exact tool_name values from the Gate 2 listing — names are not uniformly styled, some underscored, some hyphenated. Do not invent a tool name. On UNKNOWN_TOOL, read the valid names from the error and retry. Never call a tool whose effective_policy is disabled.

This map is built fresh every run and held only for this run. It is written out in the coverage sheet, never written back into this file.

Gate 0 — Confirm the workspace

Call list_workspaces with no arguments.

• One workspace → read the workspace name back and wait for an explicit yes.
• Two or more (selection_required) → ask in chat whether this is single- or multi-workspace, then which workspace(s) by name, then call again with workspace_ids=[…] and mode="single" / mode="multi".

Never auto-pick. Never print an internal numeric tenant id — name the workspace, pass the opaque ws_… handle.

Gate 1 — Discover live datasources and settle the entity scenario

Call get_agent_datasources with the confirmed workspace_id.

• connected: true → in scope.
• connected: false → excluded, and named as excluded. Surface any reconnect_url.

Settle the entity scenario:

• Single-entity — ask which company by display_name; the workflow runs against that one data_source_id.
• Multi-entity — ask which set, and confirm which entity each batch of transactions belongs to before coding any of them. Coding is entity-specific: the same vendor may map to different accounts in different company files, because the charts differ. See the cross-entity step.

Refer to companies by display_name; never show the raw data_source_id.

PART A — Explore the confirmed sources, and personalise this run

The workspace and its data sources are settled. This part finds out what they expose and which of it serves this request — the tool catalogue in Gate 2, then what is already known about this entity plus whatever still has to be asked in Gate 3. The result is a run shaped around these books, not a generic template.

Gate 2 — Discover enabled tools → build the capability map

Write tools are out of scope — always

get_datasource_tools describes what the connection could do. This skill uses only the reads.

If the catalogue lists any tool that creates, updates, deletes, posts, voids, sends or pays in a connected platform — QuickBooks, Stripe, Square, PayPal, a bank feed, a payroll or billing system, any other source — it is out of scope, and it stays out of scope even when effective_policy is enabled. A permission to write is not an instruction to write. Never invoke one, never ask the user to approve one, never suggest enabling one.

This holds for every connected platform, not only the books. Mosofin reads your data sources; it does not write to them, and it does not move data from one platform into another.

If a step appears to need a write, that step is [manual]. Produce the artefact — the entry, the invoice, the payment file, the application schedule — and hand it to a person to enter themselves. Say so plainly in the output, so nobody assumes it was done.

Hard stop — the four ways a write could slip through

Never route around this rule. Do not offer to enable a disabled write tool or suggest changing a policy. Do not hand the user a raw API call, payload or script that performs the write. Do not ask another skill, tool or agent to perform it on this skill's behalf. Do not defer it to a later step in the hope it becomes permitted.

There is no path through this skill that ends in changed data. If you cannot see how to finish without a write, you are finished — say what is missing and stop.

For each in-scope datasource (and per company file when several are live — pass data_source_id), call get_datasource_tools. Bucket every tool by effective_policy:

Resolve every task in Part B against these buckets. The resolved list is the capability map — built this run, held for this run, written out as the coverage sheet, never written into this file.

Rules that bite hardest here:

• Read the real tool name from the catalog, never from memory. Names are not uniformly styled — some underscored, some hyphenated.
• A near-substitute is not a substitute:
  • Coding history is not correctness. The central caution above.
  • A vendor's default account on its master record is not how it has actually been coded — defaults are set once and then overridden transaction by transaction. Read the history, not the default.
  • A tax code applied previously is not proof of recoverability. The code is a setting.
  • Two matching amounts are not a transfer until both legs are seen in the entity's own accounts.
• If the COA read is disabled, the original's fallbacks apply in full — category names only, or an explicitly authorized placeholder labelled PLACEHOLDER — REMAP REQUIRED BEFORE POSTING. Say which fallback is in use and why.

Gate 3 — Profile the entity, then interview the user

Call the platform's company-profile tool (on QuickBooks, get_company_info) for each in-scope entity.

Derive silently what the profile answers: legal name, base currency, fiscal calendar, country / region — the jurisdiction — and industry.

Ask the user what actually changes the work — the original Inputs table, minus the four the connected books now answer:

If no chart of accounts is available, the original's rule stands: stop and ask. Do not invent account numbers. Two acceptable fallbacks:

1. The user provides a category list (description only) and coding is to category names with no numbers.
2. The user explicitly authorizes a placeholder COA — output clearly labelled "PLACEHOLDER — REMAP REQUIRED BEFORE POSTING".

If no jurisdiction is available and the user asks for tax codes, ask. Do not assume.

Ask as one short batch. Propose defaults where reasonable — but never default a capitalization threshold or a tax treatment.

On later runs, read stored preferences first (Step 6), confirm in one line, and ask only what changed. See the Learning Loop below — it and the evolution step are the same mechanism.

PART B — The domain work

Every step below is the original procedure, unchanged in count, order, or substance, with plain-language wording, an [auto] / [gated] / [manual] verdict, and the typical evidence tool added.

Never drop a task because no tool covers it. The capitalization threshold and the tax rules are legitimately [manual].

Tool names in italics are typical. Resolve real names and policies from your Gate 2 catalog.

Step 0 — Fetch the evidence (grounding) — Mosofin addition

Batch independent reads into one message — the accounts, the vendors, the history and the tax codes do not depend on each other. Never serialize them.

The server is stateless: pass data_source_id on every call, including retries.

Typical opening batch, per in-scope entity:

• search_accounts — the real chart of accounts, including uncategorized and suspense accounts — usually [auto]
• search_vendors + get_vendor — the vendor master and any default accounts — usually [auto]
• get_general_ledger — the coding history, which is the Tier 1 evidence — usually [auto]
• search_purchases / search_bills / search_expenses — prior transactions by vendor with their accounts — usually [auto]
• get_general_ledger on Uncategorized Expense / Income and any suspense account — the work waiting to be done — usually [auto]
• search_tax_codes / search_tax_rates — the codes configured in this entity — often [auto]
• search_accounts on bank and credit card accounts — for transfer detection — usually [auto]
• search_employees — for related-party name comparison — usually [auto]
• get_company_info — legal name, currency, country, industry — usually [auto]

Handle the envelopes:

• approval_required → ask the user in chat, then re-invoke the same tool with approved=true.
• entity_required → ask by display_name, then pass that data_source_id.
• tool_policy_disabled → convert that task to [manual] and record the gap. If it is the COA read, say explicitly which fallback is now in use.
• UNKNOWN_TOOL → read the valid names from the error; do not guess.
• Dead connection → surface the reconnect_url.

Check the mock flag. mock: true is fixture data — coding history from fixtures is not this entity's history, and a Tier 1 match against it is worthless. Downgrade every history-based confidence to Low and say why.

Step 1 — Parse and normalize

For each transaction extract — [gated]: [auto] for transactions already in the system, [manual] for a pasted export:

• Date
• Vendor / payee — normalize: **strip corporate suffixes ("INC", "LLC", "LTD", "GMBH", "PTY"), trailing numeric IDs, payment-processor prefixes ("SQ ", "TST", "POS DEBIT", "PAYPAL ", "WISE ")
• Description / memo
• Amount (signed)
• Source account (bank, credit card, etc.) — if provided

Normalize examples:

• "AMZN MKTP US*1A2B3C" → "Amazon"
• **"SQ COFFEE SHOP" → "Coffee Shop (via Square)"*
• **"PAYPAL VENDORNAME 12345" → "VendorName (via PayPal)"*

Mosofin addition — match the normalized name to the vendor master. [auto] via search_vendors. Three outcomes, each worth reporting differently:

• Matches an existing vendor → the full coding history is available. Best case.
• Matches nothing → a new vendor, so Tier 1 cannot apply. Fall to Tier 2.
• Matches several similar vendors → a duplicate vendor record, which is a data-quality finding in its own right and means the history is split across records. Report it; see duplicate-invoice-detection for the matching discipline.

Step 2 — Classify each transaction

Apply this decision hierarchy in order. Stop at the first confident match.

Tier 1 — Exact vendor match (Confidence: High) — [auto]

If the vendor exists in the user's prior mappings, use that account. Reasoning: "Matches prior coding for [vendor]."

In a Mosofin workspace this tier is read, not supplied. Query the ledger history for this vendor (get_general_ledger, search_purchases) and report the account, the count and the recency:

"6120 Office Supplies — this entity's coding for this vendor in 43 of 45 prior transactions, most recently 2026-07-14."

Three history shapes, three different answers:

Frequency is evidence of practice, not of correctness. Where history conflicts with a supplied coding policy, the policy wins and the conflict is reported.

Tier 2 — Vendor-type pattern match (Confidence: High) — [gated]

Map common vendor types to the closest matching account in the user's COA. Never output an account that isn't in the COA the user provided.

If the user's COA has no matching account for a vendor type, flag it and ask the user which account to use. [auto] to determine — the real chart is readable, so whether a matching account exists is a fact here, not a question. Where none exists, say so precisely: "no Hosting or Cloud Services account exists in this chart; the nearest is 6300 Software Subscriptions — confirm."

Tier 3 — Keyword inference (Confidence: Medium) — [auto]

Tier 4 — Amount heuristics (Confidence: Medium-Low) — [auto]

• Recurring same-amount monthly charges → likely subscription. Mosofin note: the full history is available, so recurrence is a measured fact — "same amount on the same day for 11 consecutive months" — rather than a guess from one export.
• Round-number month-end amounts → likely rent, payroll, or owner draw
• Many small food-vendor charges → Meals
• Amounts ≥ entity capitalization threshold → flag as possible Fixed Asset — ask the user for the threshold if unknown; do not assume

Tier 5 — Cannot classify (Confidence: Low)

Mark as "Needs Review" with reason. Do not guess.

Step 3 — Tax codes (jurisdiction-driven)

Only suggest tax codes if the user provided a jurisdiction AND a tax code list. [gated] — the jurisdiction is now [auto] from the profile and the configured codes are [auto], but which code applies remains [manual].

Common jurisdiction patterns recognized structurally, without baking in rates or rules:

• Single-rate VAT / GST — standard / zero / exempt / out-of-scope
• Multi-rate VAT with reduced rates
• Federal + sub-national stacked indirect taxes
• US-style sales tax — paid by buyer at POS, generally not input-creditable
• No indirect tax for the entity — small supplier exemption, free zone, etc.

For each transaction suggest a tax code based on:

• Vendor's tax registration status (if known)
• Nature of supply — goods / services, domestic / cross-border
• Recoverability under the entity's regime

Cross-border / reverse charge: if a vendor appears outside the entity's jurisdiction, flag for possible reverse-charge or import-VAT treatment. Surface for review — do not silently apply. [auto] support: the vendor's address is readable, and the entity's country is readable, so the cross-border flag itself is determinable even though the treatment is not.

If no tax code list is available, output "Tax code: TBD — confirm with tax setup." Mosofin note: where the platform exposes the entity's configured codes, name the codes that exist rather than only "TBD" — it turns a blank into a short list to choose from.

Step 4 — Special-case rules

Capitalization threshold check — [gated]

For equipment, computers, furniture, vehicles and so on, if the amount ≥ the entity's capitalization threshold, flag:

"Possible capital expenditure — consider Fixed Asset. Capitalization threshold: [user-provided or 'unknown — please confirm']."

Do not pick a threshold. Ask the user if not provided. See fixed-asset-register-and-depreciation, whose threshold scan is the same test run in the other direction.

Personal-use detection — [auto] on the patterns

Flag patterns suggesting personal rather than business use:

• Groceries, personal retail
• Travel timing inconsistent with business — weekend, vacation destinations
• Personal services — gym, spa, grooming
• Family-member-named bookings

Surface as: "Possible personal — confirm business purpose." Surface, never conclude — and where a name match against the employee or owner list drives the flag, say that is what drove it.

Split-transaction detection — [auto]

Single transactions likely covering multiple categories — e.g. a warehouse-club run with mixed items — flag for split coding.

Inter-account transfer detection — [auto], and now genuinely determinable

Same amount, opposite sign, within ~3 days, between the entity's own accounts → flag as transfer, not P&L. With ledger access both legs are visible, so this moves from inference to observation: name the matching transaction and its account. The entity's own bank and card accounts are readable (search_accounts), which is what makes "between the entity's own accounts" checkable at all.

Owner / related-party transactions — [gated]

Payments to / from owners, shareholders, directors, or related entities → flag for Due to / from accounts or distributions. [auto] support: compare vendor names against the employee list and against the other connected company files' legal names. A vendor whose name matches another entity in the group is an intercompany transaction, not a third-party expense.

Loan payments — [gated]

Always split between Interest Expense (P&L) and Principal (BS liability reduction). Ask for the split if not obvious from the source. [auto] support: the loan liability account and its movement are readable, which often reveals the principal portion directly — but the amortization schedule is [manual] (see debt-and-covenant-tracking).

Step 5 — Output

Pasted lists → markdown table inline. Uploaded files → .xlsx.

| Date | Vendor (Cleaned) | Description | Amount | Suggested GL Account | Tax Code | Confidence | Reasoning | Flags |

Add two Mosofin columns: Evidence — the history count behind a Tier 1 match, or the tier used — and Verdict ([auto] / [gated] / [manual]).

Confidence:

• 🟢 High — exact vendor match or unambiguous pattern
• 🟡 Medium — keyword or amount-based inference
• 🔴 Low — needs review

Flags:

• 🏛️ Possible CapEx
• 👤 Possible personal
• ✂️ Possible split
• 🔁 Possible transfer
• ⚠️ Unusual amount
• 🌍 Cross-border / reverse-charge candidate
• 🔗 Related-party

Two Mosofin flags added:

• 📊 Inconsistent history — this vendor has been coded to more than one account
• 👥 Duplicate vendor record — the history is split across records

End with summary: total coded, % by confidence band, count of flagged items. Plus the Mosofin header block: workspace name; the entity by display_name; each excluded company file and why; whether any read returned mock data; and whether the COA was read or a fallback is in use.

For an .xlsx, add two sheets:

Sheet: Uncoded Population — NEW, Mosofin-specific — what is sitting in uncategorized and suspense accounts, with ages and amounts. This is the work list, and it is often more useful than the batch that prompted the request.

Sheet: Coverage — NEW, Mosofin-specific

| Task | Entity (display_name) | Verdict | Tool used / external source | As-at date | Policy | mock | Gap |

If creating xlsx, read first: /mnt/skills/public/xlsx/SKILL.md

Output naming: [EntityName]_Coded_Transactions_[YYYY-MM-DD].xlsx

[EntityName] is the company

…(truncated)