Tier 2: Extended
Remember Finances
Import financial documents — CSV/PDF bank statements, receipt images, invoice PDFs — and extract structured transactions with amounts, dates, merchants, and categories. Preserves the original document as a source for audit provenance.
Skill instructions
Import financial documents — bank statements, receipts, invoices — into Neotoma memory with structured transaction extraction and full provenance. When the user has financial documents (CSV/PDF bank statements, receipt images, invoice PDFs) and wants to persist the transactions as structured, queryable records. Run the ensure-neotoma skill first if Neotoma is not yet installed or configured in your current harness. | Source | Format | Method | |--------|--------|--------| | Bank statements | CSV, PDF | File read, parse file for PDFs | | Receipts |…
Preview from repository markdown.
Install and runtime steps for Remember Finances live in the file below.Agents should fetch the raw URL; humans can use GitHub browse. Follow prerequisites, MCP setup, and workflow there after neotoma setup when the skill is installed via the package.
- Repository path
skills/remember-finances/SKILL.md
Suggested prompt
Use skill prompt
Paste into your harness chat so the agent fetches the canonical markdown and follows it.
Use the Neotoma skill "Remember Finances". Fetch and follow the instructions at:
https://raw.githubusercontent.com/markmhendrickson/neotoma/main/skills/remember-finances/SKILL.md
Execute every prerequisite, MCP configuration step, and workflow phase described there for this environment.Supported data sources
Entity types produced
transactionreceiptinvoicecontactWorkflow
- 0
Verify Neotoma
Confirm MCP connectivity
- 1
Identify documents
Locate financial files and detect format
- 2
Parse and preview
Show transaction count, date range, totals
- 3
Extract transactions
Structured extraction with consistent entity_type
- 4
Store with provenance
Preserve raw document as source