KSeF Accountant (Polish)
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If you configure a real KSeF token, the agent or platform may have access to sensitive tax and invoice operations depending on the surrounding tool permissions.
The skill documents optional use of a KSeF API token, which could authorize access to invoice data and invoice operations. This is expected for a KSeF assistant and is explicitly marked optional/secret, but it is sensitive authority.
KSEF_TOKEN: description: "Token API KSeF do uwierzytelniania sesji..."
Only configure KSeF credentials through a trusted secret mechanism, prefer DEMO first, use least-privilege tokens, and do not paste tokens into chat.
If implemented or used with tools, these workflows can create legally significant tax records.
The reference material covers invoice submission to KSeF, including production use where invoices are legally binding. This is aligned with the accounting purpose, and the artifact also warns against production testing.
POST /api/online/Invoice/Send ... PRODUKCJA ... Dane: Prawne wiążące
Require explicit user confirmation before any production KSeF submission, validate invoice XML, and test only against the DEMO environment until reviewed.
A platform that does not recognize the secret declarations might not protect configured credentials as expected.
The supplied registry metadata does not declare environment variables, while SKILL.md describes optional KSEF_TOKEN, KSEF_ENCRYPTION_KEY, and KSEF_BASE_URL handling. This appears to be an under-declared metadata/registry gap rather than hidden behavior, because SKILL.md openly warns users to verify platform enforcement.
Env var declarations: none
Before adding any credentials, verify in the registry/platform UI that the variables are registered as secrets and that model invocation restrictions are enforced.
If a user implements these patterns, sensitive invoice history could influence future classifications and needs protection from poisoning or over-trust.
The AI reference patterns use historical invoices and contractor data for classification, fraud detection, and cash-flow prediction. This is purpose-aligned and described as conceptual, but invoice history is sensitive business data.
Trenowanie na danych historycznych ... historical_invoices
Keep human review for accounting decisions, protect historical invoice data, and separate training/test data from production workflows.