Financial Reporting and Forecasting Agent

1. The Problem

Board-quality financial reporting requires a finance team to compile data from accounting software, banking systems, payroll, and operational dashboards — then build a coherent narrative around the numbers that non-financial stakeholders can act on. At smaller businesses, this work falls on an already-stretched finance director or founder. At larger ones, it consumes two to three days of a CFO's time each month. The output is backward-looking by the time it lands. The decisions that needed those numbers were made last week.

2. Integration & Webhook Setup

Follow the exact steps below to configure and deploy this automation inside your OpenClaw workspace.

1. 1Create a new agent named `financial-reporting`. Connect your accounting system (Xero, QuickBooks, or Sage — all supported). The agent requires read-only API access to P&L, balance sheet, cashflow statement, and trial balance. Write access is never required for reporting.
2. 2Configure your `BUDGET_IMPORT`: upload your approved annual budget as a CSV using a standardised budget template. The agent references this for all variance analysis. Update the budget import at the start of each financial year or whenever a board-approved reforecast is issued.
3. 3Set up your `FORECAST_MODEL` parameters: define your key revenue drivers, cost assumptions, and the three scenario inputs for base, upside, and downside cases. The agent regenerates the 12-month forecast monthly, incorporating actuals to date as the anchor for forward projections.
4. 4Configure your `BOARD_PACK_TEMPLATE`: define the sections, page order, and formatting for your management accounts. OpenClaw documentation contains templates for three common board pack formats (standard SME, PE-backed, listed company). Select the closest match and adjust.
5. 5Set your distribution workflow: the completed draft pack is delivered to the finance director or CFO for review before board distribution. Configure a two-stage delivery: draft to reviewer at T+1 day, final to board at T+3 days after month end.

3. The Context Payload (context.json)

Save this file as: .openclaw/agents/financial-reporting-and-forecasting-agent/context.json

{
  "automation_id": "23",
  "title": "Financial Reporting and Forecasting Agent",
  "level": 3,
  "tier": "Enterprise Playbook",
  "setup_time": "1 week",
  "estimated_api_cost": "~$60–$120/mo",
  "client_price_range": "$2,000–$10,000",
  "agents": [
    {
      "role": "orchestrator",
      "model": "claude-3-5-sonnet",
      "temperature": 0.2,
      "max_tokens": 4096
    }
  ],
  "memory": "session",
  "output_format": "structured_json",
  "human_review_gate": true,
  "documentation_standard": "required"
}

4. Execution Commands

Run these commands from your openclaw-workshop/ directory to validate, test, and schedule this automation. Commands are taken directly from The OpenClaw Income Engine, Appendix G.

# ── STEP 1: Test accounting system connection (read-only) ──
$ openclaw run financial-reporting --test-source xero --verbose
$ 
# ── STEP 2: Upload and validate budget CSV ──
$ openclaw run financial-reporting --validate-budget ./budget_2025.csv
# Fix any account code mapping errors before first run.
$ 
# ── STEP 3: Generate prior month report (dry run — no distribution) ──
$ openclaw run financial-reporting --month last --dry-run
$ 
# ── STEP 4: Review draft board pack ──
# Draft is emailed to FD_EMAIL for review. Review, annotate, approve.
$ 
# ── STEP 5: Approve and release to board ──
$ openclaw approve financial-reporting --pack-id <id> --action release-board
$ 
# ── STEP 6: Schedule monthly automation ──
$ openclaw schedule financial-reporting --cron "0 6 1 * *"
$ 
# ── Generate a specific month's report on demand ──
$ openclaw run financial-reporting --month 2025-03
$ 
# ── Update budget mid-year (board-approved reforecast) ──
$ openclaw run financial-reporting --update-budget ./reforecast_q3_2025.csv
$ 
# ── Run scenario model only (no full pack) ──
$ openclaw run financial-reporting --scenarios-only --month last