template-instantiator
Run the following command in DeepSeek Harness:
dsh plugin install ChenneyZhuang/template-instantiator
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install ChenneyZhuang/template-instantiator in your terminal to install; the source repository is https://github.com/ChenneyZhuang/template-instantiator .
About this plugin
The most telling trace that an agent produced a file is not a typo. It is the John Doe still sitting in row 3, the Acme Corp in the footer, the previous quarter numbers left in the update-this cell. Templates carry two kinds of residue: the visible placeholders ({{var}}, [X], lorem ipsum) and the example content you stop noticing after the tenth read. template-instantiator turns template filling into a mechanical four-step loop: inventory every placeholder and variation, replace strictly from user-provided data (a missing value becomes a question, never an invention; unrequested optional sections are deleted, not half-filled), scan the final file against the full inventory list, and record the result in the delivery note. Zero residue is not a claim; it is a shown scan.
Who it helps: teams and individuals who have agents produce reports, proposals, contract drafts, or client deliverables at scale. You do not trust that the agent remembered to clean everything; you trust the loop — inventory, replace, scan, show the zero. It does not judge whether a legal clause is correct, but it ensures no sample residue slips through. The more complex and versioned your template is, the more value this mechanical pass delivers. The one boundary: the scan only finds patterns that were inventoried; a placeholder style never listed escapes. The inventory step is the whole game.
Use Cases
- Auto-strip sample data and placeholders from agent-generated client proposals
- Pre-fill contract templates with zero residue from prior versions
- Run a mechanical full-pattern scan before delivering templated reports
Best For
- Ops or BD teams producing client deliverables at scale via agents
- Legal and business staff repeatedly filling contract and proposal templates
- Tech leads who hold agent output to a zero-residue standard
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.