Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/acdeveloper-sci/fortriage/llms.txt

Use this file to discover all available pages before exploring further.

ForTriage ships with a small, pre-packaged Fortran corpus that you can load with a single click. This is the fastest way to explore all dashboard views and AI Insights without having a Fortran project available.
Click Load sample fixtures, then click Run analysis to see ForTriage’s full output in under 60 seconds.

What is the sample corpus?

The sample corpus is a compact but representative set of Fortran sources bundled directly with the ForTriage repository at tests/fixtures/fixtures.zip. It is the same corpus used by Forti4D’s own test suite, so it exercises the full pipeline end-to-end. At a glance:
  • 26 program units across 8 source files
  • Demonstrates a variety of risk tiers, coupling patterns, and dead code scenarios
  • Small enough to analyze in seconds, but complete enough to populate every dashboard view — tier cards, scatter chart, clone report, COMMON block coupling, and AI Insights
The sample fixtures are a demo aid — they do not represent production-scale Fortran code. Real enterprise Fortran projects can contain thousands of units spread across hundreds of files.

How to load the sample corpus

1

Open the Overview tab

Launch ForTriage in your browser. The Overview tab is shown by default on every fresh session.
2

Click Load sample fixtures

Below the file uploaders, click 📦 Load sample fixtures. The caption beneath the button changes to:
“Sample fixtures loaded — click “Clear all” to remove them.”
No file upload dialog opens — the fixtures are loaded directly from the repository bundle.
3

Click Run analysis

Click Run analysis. The pipeline processes the sample corpus and populates every tab of the dashboard.
You can replace the sample fixtures with your own code at any time. Uploading a file directly or via ZIP automatically takes precedence over the fixtures. To fully reset the upload area, click 🗑️ Clear all.

What you will see

After analysis completes, the Overview tab shows the following results for the sample corpus:
MetricValue
Units analyzed26
Dead code units17
COMMON block couplingSHARED_DAT block across 2 units and 2 files
Clone pairs detected6
Here is what each result means for a real modernization project: 17 dead code units (unreachable). More than half of the sample corpus is flagged as unreachable from any known entry point. In a production codebase this is a strong signal to audit reachability against your actual entry points (main programs, external call sites, test harnesses) before scheduling refactoring effort — removing live code that was incorrectly classified would be costly. COMMON block coupling (SHARED_DAT). Two units in two different files share state through a COMMON block. This kind of implicit global coupling is one of the primary modernization hazards in legacy Fortran — any change to the block’s layout silently affects every unit that references it. The Coupling tab in ForTriage maps the full extent of this coupling across your real codebase. 6 clone pairs. Six pairs of near-duplicate program units were detected. Clone pairs often indicate copy-paste reuse that diverged over time. In a modernization project, each clone pair is a candidate for consolidation into a shared subroutine or module — reducing the surface area for future bugs.

Using sample fixtures with AI Insights

After running analysis on the fixtures, AI Insights automatically generates a structured risk summary — no extra steps required. The summary covers high-priority units, coupling risks, and clone patterns drawn from the sample corpus. The AI Insights chat is pre-loaded with context about the sample results and can answer follow-up questions, for example:
  • “Which units should I prioritize for refactoring?”
  • “What is the risk associated with the SHARED_DAT COMMON block?”
  • “Explain why so many units are flagged as dead code.”
AI Insights requires a Gemini API key. See AI Insights setup for configuration instructions.

Build docs developers (and LLMs) love