ArduPilot · github-actions · Dec 1, 2025 · Dec 1, 2025 · Dec 2, 2025 · Dec 3, 2025
@@ -8,3 +8,6 @@ messages:
       Please read the translation file "{{translation_file}}" and translate all the strings from English to {{language}}.
 
       The file contains strings in the format "line_number:English text" - please translate only the text after the colon while preserving the exact line number and colon format.
+
+      Translation strings to process:
+      {{translation_content}}
@@ -207,7 +207,7 @@ jobs:
       matrix:
         include: ${{ fromJson(needs.extract_strings.outputs.translation-matrix) }}
       fail-fast: false  # Continue processing other languages even if one fails
-      max-parallel: 5   # Limit concurrent AI requests
+      max-parallel: 1   # Limit concurrent AI requests to avoid rate limiting
 
     steps:
       - name: Harden the runner (Audit all outbound calls)
@@ -236,6 +236,15 @@ jobs:
           echo "📊 Translation file size: $file_size bytes"
           echo "✅ Translation file validation completed successfully"
 
+      - name: Prepare translation content
+        id: translation_content
+        run: |
+          # Read file and indent each line by 2 spaces (except the first)
+          awk 'NR==1 {print} NR>1 {print "  " $0}' "${{ matrix.file }}" > indented_content.txt
+          echo "content<<EOF" >> $GITHUB_OUTPUT
+          cat indented_content.txt >> $GITHUB_OUTPUT
+          echo "EOF" >> $GITHUB_OUTPUT
+
       - name: Run AI translation
         id: ai_translate
         uses: actions/ai-inference@334892bb203895caaed82ec52d23c1ed9385151e # v2.0.4
@@ -246,10 +255,8 @@ jobs:
             language: ${{ matrix.language }}
             lang_code: ${{ matrix.lang_code }}
             translation_file: ${{ matrix.file }}
-          file_input: |
-            translation_content: ${{ matrix.file }}
-          enable-github-mcp: true
-          github-mcp-toolsets: "context,repos"
+            translation_content: |
+              ${{ steps.translation_content.outputs.content }}
           model: openai/gpt-4.1
           max-tokens: 8000
           token: ${{ secrets.AMC_COPILOT_TOKEN_CLASSIC }}
@@ -308,22 +315,32 @@ jobs:
           # Save the AI response back to the original translation file
           if [ "${{ steps.check_translation.outputs.output_method }}" = "file" ]; then
             echo "📄 Using response file: ${{ steps.ai_translate.outputs.response-file }}"
-            cp "${{ steps.ai_translate.outputs.response-file }}" "${{ matrix.file }}"
+            cp "${{ steps.ai_translate.outputs.response-file }}" "${{ matrix.file }}.raw"
           elif [ "${{ steps.check_translation.outputs.output_method }}" = "content" ]; then
             echo "📝 Using response content"
-            echo "${{ steps.ai_translate.outputs.response }}" > "${{ matrix.file }}"
+            echo "${{ steps.ai_translate.outputs.response }}" > "${{ matrix.file }}.raw"
           else
             echo "❌ Unexpected output method: ${{ steps.check_translation.outputs.output_method }}"
             exit 1
           fi
 
-          # Validate the saved file
+          # Clean up AI output: keep only lines starting with number followed by colon
+          echo "🧹 Cleaning AI output..."
+          grep -E '^[0-9]+:' "${{ matrix.file }}.raw" > "${{ matrix.file }}" || {
+            echo "❌ Failed to extract valid translations from AI output"
+            echo "Raw output preview:"
+            head -20 "${{ matrix.file }}.raw"
+            exit 1
+          }
+
+          # Validate the cleaned file
           if [ -f "${{ matrix.file }}" ] && [ -s "${{ matrix.file }}" ]; then
-            echo "✅ AI translation saved successfully for ${{ matrix.language }} (${{ matrix.file }})"
-            echo "📊 File size: $(wc -c < "${{ matrix.file }}") bytes"
-            echo "📊 Line count: $(wc -l < "${{ matrix.file }}") lines"
+            echo "✅ AI translation saved and cleaned successfully for ${{ matrix.language }}"
+            echo "📊 Raw file size: $(wc -c < "${{ matrix.file }}.raw") bytes"
+            echo "📊 Cleaned file size: $(wc -c < "${{ matrix.file }}") bytes"
+            echo "📊 Valid translation lines: $(wc -l < "${{ matrix.file }}") lines"
           else
-            echo "❌ Translation file is empty or missing: ${{ matrix.file }}"
+            echo "❌ Translation file is empty or missing after cleanup: ${{ matrix.file }}"
             exit 1
           fi
 
@@ -437,7 +454,10 @@ jobs:
 
       - name: Insert AI translations into .po files
         if: needs.extract_strings.outputs.translations-to-process == 'true'
+        shell: bash  # Don't use -e flag to prevent premature exit
         run: |
+          set -x  # Enable command tracing for debugging
+
           # Check if we have any translated files
           if ls missing_translations_*.txt 1> /dev/null 2>&1; then
             echo "📥 Processing AI translations..."
@@ -450,12 +470,12 @@ jobs:
               if [ -f "$file" ]; then
                 if grep -q "# Translation failed" "$file" 2>/dev/null; then
                   echo "⚠️ Found failed translation: $file"
-                  ((failed_translations++))
+                  failed_translations=$((failed_translations + 1))
                   # Remove failed translation files so they don't get processed
                   rm "$file"
                 else
                   echo "✅ Found successful translation: $file"
-                  ((successful_translations++))
+                  successful_translations=$((successful_translations + 1))
                 fi
               fi
             done
@@ -466,8 +486,39 @@ jobs:
 
             if [ $successful_translations -gt 0 ]; then
               echo "🔄 Processing successful translations with insert_missing_translations.py"
-              python insert_missing_translations.py
-              echo "✅ AI translations inserted into .po files"
+
+              # Show files that will be processed
+              echo "Files to process:"
+              ls -lh missing_translations_*.txt
+
+              # Validate file format before processing
+              echo "Validating translation file format..."
+              for file in missing_translations_*.txt; do
+                echo "Checking $file:"
+                if grep -qE '^[0-9]+:' "$file"; then
+                  echo "✅ File format is valid"
+                else
+                  echo "❌ ERROR: File $file does not contain valid translation lines (format: number:text)"
+                  echo "File contents:"
+                  cat "$file"
+                  exit 1
+                fi
+
+                # Show file preview
+                echo "First 5 lines of $file:"
+                head -5 "$file"
+                echo "---"
+              done
+
+              # Run with full error output captured
+              echo "Running insert_missing_translations.py..."
+              if python insert_missing_translations.py 2>&1; then
+                echo "✅ AI translations inserted into .po files"
+              else
+                exit_code=$?
+                echo "❌ insert_missing_translations.py failed with exit code $exit_code"
+                exit $exit_code
+              fi
             else
               echo "⚠️ No successful translations to process"
             fi
@@ -536,14 +587,7 @@ jobs:
 
             🤖 **AI-Powered Translation Applied with Enhanced Matrix Processing**:
             - Automatically extracted missing translations using `extract_missing_translations.py`
-            - Used GitHub Actions matrix strategy to process numbered files in parallel
             - Applied AI-powered translations using GitHub Models (GPT-4o) for multiple languages
-            - **GITHUB PROMPT.YML FORMAT**: Using official GitHub prompt.yml template format with separated files
-            - **SEPARATED PROMPT.YML FILES**: Organized in .github/prompts/ directory for better structure
-            - **CLEAN ORGANIZATION**: ai-translation-system.prompt.yml and ai-translation-user.prompt.yml separated from workflows
-            - **PERSONAL ACCESS TOKEN**: Using amilcarlucas PAT for GitHub MCP access instead of GITHUB_TOKEN
-            - **GITHUB MCP ENABLED**: AI can read translation files directly from repository using Model Context Protocol
-            - **FILE-BASED PROMPTS**: AI reads translation files directly instead of embedding content in YAML prompts
             - Supports processing unlimited translations per language with automatic chunking
             - Inserted translated strings into .po files using `insert_missing_translations.py`
             - Compiled binary .mo files for immediate use
@@ -553,23 +597,12 @@ jobs:
             **Languages processed**: Portuguese (pt), German (de), Italian (it), Japanese (ja), Chinese Simplified (zh_CN)
 
             **Enhanced Matrix Processing & Scaling**:
-            - ✅ **Parallel processing** of translation files for better performance
             - ✅ **Automatic chunking** when >50 strings per language (configurable)
             - ✅ **Robust error handling** for failed AI translation requests with detailed debugging
             - ✅ **File validation** before and after AI processing
             - ✅ **Consistent terminology** guidelines applied across all chunks for each language
             - Robust error handling for failed AI translation requests
 
-            **Technical Improvements Made**:
-            - 🔧 **Organized prompt structure**: Moved prompt files to .github/prompts/ directory to avoid confusion with workflows
-            - 🔧 **GitHub MCP enabled**: AI can read translation files directly from repository using Model Context Protocol
-            - 🔧 **Separated prompt architecture**: System prompt and user prompt in separate files for better maintainability
-            - 🔧 **File-based AI prompts**: AI reads translation files directly, eliminating YAML content embedding issues
-            - 🔧 **Reusable system prompts**: System prompt can be reused across different translation tasks
-            - 🔧 **Enhanced reliability**: No more YAML syntax issues from embedded content with special characters
-            - 🔧 **Better scalability**: File-based approach handles large translation batches without prompt size limits
-            - 🔧 **403 error fix**: Enabled GitHub MCP to resolve permission issues when reading repository files
-
             **Translation Guidelines Applied**:
             - Technical aviation/drone context preservation
             - Formal register for technical documentation

@@ -37,7 +37,7 @@ repos:
           - tomli
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.14.6
+    rev: v0.14.7
     hooks:
       - id: ruff-check
         args: [ --fix ]
@@ -56,7 +56,7 @@ repos:
   #      entry: markdown-link-check
 
   - repo: https://github.com/igorshubovych/markdownlint-cli
-    rev: v0.45.0
+    rev: v0.46.0
     hooks:
       - id: markdownlint
         args: ['--config', '.markdownlint.yaml']
@@ -94,7 +94,7 @@ repos:
       - id: reuse-lint-file
 
   - repo: https://github.com/gitleaks/gitleaks
-    rev: v8.29.1
+    rev: v8.30.0
     hooks:
       - id: gitleaks
 

@@ -81,14 +81,14 @@ We need contributors to help realize this roadmap! Here's how you can get involv
 - 2025.09.04 - Added [empty, unopinionated vehicle templates](https://github.com/ArduPilot/MethodicConfigurator/releases/tag/v2.1.0)
 - 2025.08.28 - Option to [reset all parameters to their defaults before starting the configuration](https://github.com/ArduPilot/MethodicConfigurator/releases/tag/v2.0.6)
 - 2025.07.16 - [YouTube beginners Tutorial](https://www.youtube.com/watch?v=tM8EznlNhgs&list=PL1oa0qoJ9W_89eMcn4x2PB6o3fyPbheA9)
-- 2025.06.15 - The tunning guide is the [single most liked post in the ArduPilot forum](https://discuss.ardupilot.org/badges/20/great-topic)
+- 2025.06.15 - The tuning guide is the [single most liked post in the ArduPilot forum](https://discuss.ardupilot.org/badges/20/great-topic)
 - 2025.04.26 - [First YouTube video Tutorial](https://www.youtube.com/watch?v=47RjQ_GarvE)
 - 2025.04.02 - Initial ArduPlane template and support
 - 2025.02.20 - [AI powered, ArduPilot trained chatbot](https://gurubase.io/g/ardupilot)
 - 2025.01.27 - [40% test coverage](https://coveralls.io/builds/71941955)
 - 2024.12.18 - Reached all initially defined project features
 - 2024.03.14 - [First public release](https://discuss.ardupilot.org/t/new-ardupilot-methodic-configurator-gui/115038)
-- 2023.12.15 - Posted the [tunning guide](https://discuss.ardupilot.org/t/how-to-methodically-configure-and-tune-any-arducopter/110842)
+- 2023.12.15 - Posted the [tuning guide](https://discuss.ardupilot.org/t/how-to-methodically-configure-and-tune-any-arducopter/110842)
 
 <!-- Gurubase Widget -->
 <script async src="https://widget.gurubase.io/widget.latest.min.js"

@@ -446,6 +446,7 @@ msgstr "Validierungsschema konnte nicht geladen werden"
 msgid "Invalid JSON file '%s': %s"
 msgstr "Ungültige JSON-Datei '%s': %s"
 
+msgstr "Validierungsfehler"
 #: ardupilot_methodic_configurator/backend_filesystem_json_with_schema.py:103
 msgid "File '%s' not found in %s."
 msgstr "Datei '%s' nicht gefunden in %s."
@@ -811,6 +812,7 @@ msgstr "Autopilot-Typ {self.info.autopilot}"
 msgid ""
 "Vehicle type: {self.info.mav_type} running {self.info.vehicle_type} firmware"
 msgstr ""
+msgstr "Verbindung mit System-ID %s, Komponenten-ID %s hergestellt."
 "Fahrzeugtyp: {self.info.mav_type}, der {self.info.vehicle_type} Firmware "
 "läuft"
 
@@ -3902,10 +3904,12 @@ msgstr "'{param_name}' wurde in der apm.pdef.xml Datei nicht gefunden."
 
 #: ardupilot_methodic_configurator/data_model_parameter_editor.py:1563
 msgid ""
+msgstr "Parameter nicht gefunden"
 "Can not add parameter when no FC is connected and no apm.pdef.xml file "
 "exists."
 msgstr ""
 "Parameter kann nicht hinzugefügt werden, wenn kein FC verbunden ist und "
+msgstr "Parameter {param_name} konnte nicht gefunden werden."
 "keine apm.pdef.xml Datei existiert."
 
 #: ardupilot_methodic_configurator/data_model_parameter_editor.py:1703
@@ -4591,6 +4595,7 @@ msgstr "Anwendungssymbol konnte nicht geladen werden: %s"
 #: ardupilot_methodic_configurator/frontend_tkinter_base_window.py:337
 msgid "Image filepath cannot be empty"
 msgstr "Bilddateipfad darf nicht leer sein"
+msgstr "Allgemeiner AMC-Workflow"
 
 #: ardupilot_methodic_configurator/frontend_tkinter_base_window.py:341
 msgid "Image height must be positive"
@@ -5195,6 +5200,7 @@ msgstr "In Sequenz testen"
 
 #: ardupilot_methodic_configurator/frontend_tkinter_motor_test.py:227
 msgid "Stop All Motors"
+msgstr "Parameter konnten nicht heruntergeladen werden"
 msgstr "Alle Motoren stoppen"
 
 #: ardupilot_methodic_configurator/frontend_tkinter_motor_test.py:230
@@ -6121,7 +6127,7 @@ msgstr "Abgeleiteter Parameter"
 #: ardupilot_methodic_configurator/frontend_tkinter_parameter_editor_table.py:516
 msgid ""
 "Parameter %s changed, will later ask if change(s) should be saved to file."
-msgstr ""
+msgstr "Unbekannter Fehler."
 "Parameter %s geändert, es wird später gefragt, ob die Änderung(en) in die "
 "Datei gespeichert werden sollen."
 
@@ -6571,43 +6577,53 @@ msgid "Format version"
 msgstr "Formatversion"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:32
+msgstr "Dies ist keine Bodenstation und hat einen anderen Workflow:"
 msgid "Frame"
 msgstr "Rahmen"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:33
+msgstr "ArduPilot Methodischer Konfigurator - Workflow"
 msgid "GNSS Receiver"
 msgstr "GNSS-Empfänger"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:35
+msgstr "[Bild nicht gefunden: AMC_general_workflow.png]"
 msgid "Manufacturer"
 msgstr "Hersteller"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:36
+msgstr "siehe"
 msgid "Model"
 msgstr "Modell"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:37
+msgstr "Schnellstartanleitung"
 msgid "Motors"
 msgstr "Motoren"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:38
 msgid "Notes"
+msgstr ","
 msgstr "Notizen"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:39
 msgid "Number of cells"
+msgstr "YouTube-Tutorials"
 msgstr "Anzahl der Zellen"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:40
 msgid "Poles"
+msgstr "Anwendungsfälle"
 msgstr "Pole"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:41
 msgid "Product"
+msgstr "Benutzerhandbuch."
 msgstr "Produkt"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:42
 msgid "Program version"
+msgstr "Ich verstehe das"
 msgstr "Programmversion"
 
 #: ardupilot_methodic_configurator/vehicle_components.py:43