delivery-checklist
Run the following command in DeepSeek Harness:
dsh plugin install ChenneyZhuang/delivery-checklist
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install ChenneyZhuang/delivery-checklist in DeepSeek Harness to install, source code at https://github.com/ChenneyZhuang/delivery-checklist
About this plugin
A single wrong number in a client-facing file costs more trust than ten correct ones earn back. Yet most delivery failures trace not to a computation bug but to a habit: trusting the script that built the file instead of opening the file itself. delivery-checklist turns that pre-ship check into a repeatable five-step routine, catching in minutes the errors that would otherwise surface days later.
The five checks are: confirm column structure and header language match what the recipient acts on; open the real file to verify row counts, deduplicate, and scan for placeholders such as TODO, XXX, or stray test values; reconcile every number in the accompanying report back to the file, diffing incremental sheets against their stated baseline so that an "all new" claim becomes a verified fact; place the file in the agreed directory and log a ledger row; then re-read the placed copy exactly as the recipient will. Each step has a concrete pass-or-fail criterion rather than a gut feel.
Built for any agent workflow that ships xlsx or csv tables to clients, downstream teams, or partners. A heuristic dedup key, a ledger entry, a baseline diff—minutes of verification that protect the full chain of credibility behind every delivery.
Use Cases
- Verify all-new claims on incremental spreadsheet deliveries
- Scan client-facing files for placeholders and duplicate rows before shipping
- Deduplicate multi-batch entity lists and reconcile the ledger
Best For
- Agent workflows that ship xlsx or csv tables to clients
- Teams managing multi-batch incremental data
- Data analysts requiring a pre-delivery quality gate
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.