aibot88 tachi-report-assembly
PDF security report assembly domain knowledge extracted from the tachi report-assembler agent. This skill provides the artifact detection tables, Typst template interface contract, and brand asset guidelines that the report-assembler uses to transform pipeline artifacts into a professional PDF security assessment booklet.
Domain Overview
The report assembly pipeline bridges tachi pipeline artifacts (markdown and image files) with a Typst template system to produce a multi-page PDF:
- Artifact Detection -- Identify which pipeline outputs are available in the target directory and select the richest data source tier for findings presentation
- Data Extraction -- Parse markdown artifacts and extract structured data via deterministic Python script
- Typst Data Generation -- Generate a
report-data.typfile containing all extracted variables as Typst#letbindings that the template system imports at compile time - Brand Asset Resolution -- Locate logo files with dark variant support and compute relative paths for Typst image inclusion
The Typst template system uses a data injection pattern: main.typ imports report-data.typ and conditionally includes page templates based on artifact availability flags. The agent's responsibility is generating report-data.typ with correct variable names, types, and values.
Loading Table
Reference files are loaded on-demand by the report-assembler agent at specific workflow phases using the Read tool.
| Reference | File | Load When |
|---|---|---|
| Typst Artifacts | references/typst-artifacts.md |
Artifact detection phase (Step 1) — determining which pipeline outputs exist and selecting data source tier |
| Typst Template Contract | references/typst-template-contract.md |
Data generation phase (Step 2) — generating report-data.typ with correct variable bindings |
| Brand Asset Guidelines | references/brand-asset-guidelines.md |
Brand detection phase (Step 2) — locating and resolving logo paths for cover and footer |