AI agents for corporate accounting on Autopilot

/loop 30m Start the "AI agents for corporate accounting on Autopilot" loop.

Inspired by FullSeam (https://fullseam.com).

Goal: open work triaged, exceptions owned, and core finance workflow current with audit trail
Max iterations: 20
Between iterations run: Report open queue items, stale tasks, failed automations, and items awaiting human approval for FullSeam
Exit when: zero open items without owner or explicit escalation, all external actions approved or sent, and systems of record current

Step 1 — Ingest transactions: Pull bank feeds, invoices, receipts, and ledger exports.
Step 2 — Code and match: AI categorizes entries and applies matching rules.
Step 3 — Investigate exceptions: Gather evidence for unmatched or ambiguous items.
Step 4 — Human approval: Queue material adjustments and period-close entries for sign-off.
Step 5 — Close and report: Export audit-ready summaries and update ERP.

## Before you start

Connect plugins:
- GitHub (required) — Read branches, PRs, reviews, checks, workflow runs, and source diffs.

Attach skills:
- Loop runner (required) — Self-pace iterations, run the check between passes, and stop only on the exit condition.
- Code change + local verification (optional) — Edit code safely, run commands, and keep changes scoped.
- CI debugging (optional) — Read failing checks, logs, and the smallest actionable root cause.
- Approval workflows (optional) — Keep outbound actions in draft or approval states when risk is non-trivial.
- Security triage (optional) — Assess alerts, prioritize patches, and avoid unsafe shortcuts.

Self-pace this loop. After each iteration, run the check command, read the output, and only continue if the exit condition is not met. Stop when the exit condition passes or max iterations is reached. Give a short status update each pass.