Skip to main content
For AI agents: a documentation index is available at https://docs.coverbase.com/llms.txt. This page is also available in markdown by appending .md to the URL.
This is the complete list of placeholders you can write in an assessment report template. Field names are matched case-insensitively, and whitespace around the entity, the ::, and the field name is ignored. The same list is in the product: Variables Catalog in the Add new assessment report template dialog shows every placeholder with its exact spelling, including your organization’s enabled custom fields, and copies one when you click it.

Vendor

{{vendor::<field>}} resolves against the vendor that the assessment belongs to.

Assessment

{{assessment::<field>}} resolves against the assessment being exported. Dates are formatted YYYY-MM-DD in the timezone supplied at export time.

Services

{{services::<field>}} covers all services in scope on the assessment, flattened into a single value. When the assessment has one service, the value is the bare field value. When it has more than one, the value becomes a bulleted block of • Service name - value, and services with no value for that field are dropped. When the assessment has no services, every services placeholder is empty.

Reviews

{{reviews::<field>}} covers the whole-assessment reviews on the assessment. Domain-scoped reviews from the risk profile are excluded, matching the other exports. Formatting follows the same rule as services: a single review renders bare, several reviews render as a bulleted block keyed by a row label. The row label is the review name, falling back to the reviewer, falling back to Review 1, Review 2, and so on.
Use {{reviews::summary}} when there is more than one reviewer. The individual field placeholders each produce their own bulleted list, so reading a reviewer’s outcome next to their notes means cross-referencing two lists. summary keeps each reviewer’s label, outcome, and notes on one line.

Vendor Intelligence

{{intelligence::<field>}} resolves against Coverbase’s outside-in research on the vendor — the data behind the Vendor Intelligence tabs — plus any financial statements your organization has submitted for that vendor. These are stored values, so they resolve exactly, without a model call. Prefer them over {{agent::...}} for any figure that is a fact rather than a judgement: a placeholder cannot get a market capitalization wrong. Every {{intelligence::...}} placeholder resolves to something. A field with no value for this vendor (a private company has no market capitalization) renders Not available rather than staying in the document as placeholder text. So does a field from a Vendor Intelligence surface your organization is not entitled to.

Company and market

Financial health

Security posture

Corporate registration

People, reputation and screening

Submitted financial statements

These resolve only when your organization has uploaded financial statements for the vendor. The metric blocks each render one line per metric, with both fiscal years and the change between them.
A figure the statements do not report renders as Not reported, never as 0. A zero would read as a measured result, which is a different claim about the vendor.

Findings

{{findings::<set>::<column>}} only works inside a table row. The row is cloned once per finding in the set. See repeating findings tables for how to lay the table out, and one table per risk domain to split findings by risk domain. Sets: Columns, usable with either set: Rows where a column has no value render that cell empty. Each set is capped at 100 rows.

Scoping a table to one risk domain

{{findings::<set>::<risk domain>::<column>}} narrows a table to the findings whose controls belong to one risk domain, so a report can carry a separate table per domain:
The domain is the name as configured in Configuration → Risk domains, matched ignoring case and surrounding spaces. It resolves per finding through the control the finding was raised against: the control section’s risk domain when it sets one, otherwise the control set’s. A domain with no findings renders “No new findings for Privacy”; a domain your org does not have is reported as an unknown placeholder.

Agent

{{agent::<your prompt>}} generates prose from the assessment. The whole text after :: is the prompt. The agent reads the assessment, its findings, evals, reviews, services and commitments, and questionnaire responses, and the headline of every Vendor Intelligence section — company profile, financial health, security posture, corporate registration, people, reputation, compliance and legal, products, and sanctions screening. It can pull any of those sections in full when a prompt needs the evidence behind a headline. Use a placeholder rather than the agent for anything that is a stored value. The agent is for judgement: a conclusion, an overview, a paragraph weighing several facts together.
Output is rendered as Word content, honoring **bold**, *italic*, ***bold italic***, ==highlight==, and bulleted and numbered lists. Headings and markdown tables are not rendered. See AI-written sections for the full behavior, including timeout and failure markers.

Custom field formatting

Custom fields render according to their configured type.
Select and multi-select fields with more than 10 options collapse to a plain comma-separated list of only the selected values. A marked list of 70 country options is unreadable in a report, so Coverbase drops the unselected ones automatically.

Quick lookup

{{vendor::name}} and {{vendor::description}}.
{{assessment::residual risk}} for the outcome, {{assessment::inherent risk}} for the starting point, {{assessment::assessment score}} for the numeric score.
A two-row table with {{findings::new::id}}, {{findings::new::title}}, {{findings::new::risk level}}, and {{findings::new::due date}} in the body row.
One table per domain, each using {{findings::new::<risk domain>::<column>}}, for example {{findings::new::Privacy::title}}. Or keep one table and add a {{findings::new::risk domain}} cell to label each row.
{{reviews::summary}} for everything on one line per reviewer, or {{assessment::approval option}} for the approval decision recorded on the assessment itself.
{{intelligence::financial health score}}, {{intelligence::financial health band}} and {{intelligence::security rating}}. For a credit memo built on submitted statements, {{intelligence::financial risk rating}} and the metric blocks.
{{agent::...}} with a specific prompt. Say who the audience is and how long the answer should be.