Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.appliedaifoundation.org/llms.txt

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

​
What it does

The Maritime Report Generator takes a folder of inspection photos plus structured vessel data and produces a finished HTML report. It runs the Image Analysis skill over the photos, maps each photo to the right section of the chosen report template, fills in vessel details, and outputs a single self-contained file ready to send.

​
Report types

  • Vessel Inspection Report (VIR) — TSI-format inspection covering hull, machinery, accommodation, safety
  • Owner’s Report — owner-facing condition summary
  • Incident Investigation Report — root-cause analysis with timeline, evidence, and corrective actions
  • Dry Dock Report — pre/post dry-dock condition with work-list verification

​
When to use it

  • “Generate vessel inspection report”
  • “Create VIR”
  • “Owner’s report for AQUILA”
  • “Incident investigation report”
  • “Dry dock report”
  • “Inspection report from these photos”

​
How it works

  1. Identify report type and vessel data — confirms the report template and gathers vessel details (name, IMO, dates, location).
  2. Analyse photos — calls the image-analysis skill in vessel mode to classify equipment, rate condition, detect defects.
  3. Map photos to sections — matches each analysed photo to the relevant report section (engine room, deck, cargo, accommodation, safety, etc.) based on equipment type.
  4. Fill the template — writes the report HTML with vessel data, photos, captions, condition summaries, and any flagged findings.
  5. Present draft — delivers a self-contained HTML file the user can review, iterate on, or send.

​
What you can control

  • Report type (VIR / Owner’s / Incident / Dry Dock)
  • Vessel data (name, IMO, inspection date, location, attendees)
  • Inclusion of executive summary, defect list, recommendations
  • Photo selection and order within each section

​
Output

A single HTML file with embedded photos and styling — works offline, sends as one attachment, no external assets.

​
Why it’s an orchestration skill

Generating a report is multi-step: photo analysis, classification, mapping, templating. Doing this skill-by-skill works but takes care to wire correctly. The generator handles the wiring so the user provides photos + a few details and gets a draft back.
  • image-analysis — runs the per-photo equipment classification this skill consumes
  • qhse-investigation-report — specialised template for safety incidents
  • pdf-to-markdown — useful when input includes scanned reports