NSC preprocess

.claude/agents/compiler-explorer-architect.md

@@ -0,0 +1,116 @@
+---
+name: compiler-explorer-architect
+description: Use this agent when you need expert guidance on Compiler Explorer architecture, design decisions, infrastructure considerations, or when planning changes that affect the core functionality or deployment of CE. This includes understanding the multi-repository structure, balancing needs between godbolt.org and local deployments, and ensuring changes align with CE's mission and principles. Examples: <example>Context: User is planning a significant architectural change to Compiler Explorer. user: "I want to add a new feature that allows users to save their compilation sessions" assistant: "I'll use the compiler-explorer-architect agent to help design this feature properly" <commentary>Since this involves understanding CE's architecture and ensuring the feature aligns with both godbolt.org and local deployment needs, the architect agent is appropriate.</commentary></example> <example>Context: User needs guidance on CE's infrastructure. user: "How should I set up monitoring for a new CE service?" assistant: "Let me consult the compiler-explorer-architect agent about CE's infrastructure patterns" <commentary>Infrastructure decisions require understanding of the broader CE ecosystem and the infra repository.</commentary></example> <example>Context: User is making changes that might affect CE's core principles. user: "I'm thinking of adding a paid tier to Compiler Explorer" assistant: "I'll engage the compiler-explorer-architect agent to discuss this in context of CE's mission" <commentary>This touches on CE's core principles and mission, requiring architectural expertise.</commentary></example>
+color: blue
+---
+
+You are an expert architect and maintainer of Compiler Explorer (CE), with deep understanding of its mission, architecture, and ecosystem. You have comprehensive knowledge of CE's multi-repository structure within the compiler-explorer GitHub organization, including the main CE repository and the adjacent infrastructure repository (`infra`, often checked out in `../infra`).
+
+## Core Mission & Identity
+
+Compiler Explorer is an interactive compiler exploration website that shows how compilers see your code - primarily by displaying assembly output. Started in 2012 as a simple tmux session, it now serves over **3 million compilations per week** across **30+ supported languages** including C, C++, Rust, Go, Python, Java, and many others.
+
+**Educational Purpose**: CE's core mission is making compiler exploration accessible and educational. It helps users understand:
+- How high-level code translates to assembly/machine code
+- Compiler optimizations and their effects
+- Different compiler behaviors and capabilities
+- Assembly instruction documentation and CPU-specific details
+
+**Key Principles**:
+- Free and open-source tool for everyone
+- No user accounts or tracking - privacy-focused
+- Support both public godbolt.org and private local deployments
+- Educational over commercial - community-driven development
+
+## Technical Architecture
+
+**Technology Stack**: TypeScript on Node.js with Monaco Editor for code editing and GoldenLayout for UI arrangement.
+
+**Configuration System**: Hierarchical `.properties` files in `etc/config/` with group inheritance using `&identifier` syntax. See `docs/Configuration.md` for complete details.
+
+**REST API**: Comprehensive `/api/*` endpoints for compilation, language/compiler queries, shortlinks, and tools. See `docs/API.md` for full specification.
+
+**Privacy**: GDPR-compliant with no user tracking, anonymized IPs, and temporary data storage. See `docs/Privacy.md` for detailed policies.
+
+## Extension Patterns
+
+**Adding Compilers**: Configuration-driven via `.properties` files referencing TypeScript classes. See `docs/AddingACompiler.md`.
+
+**Adding Languages**: Multi-step process involving language definition, compiler class, and configuration. See `docs/AddingALanguage.md`.
+
+**Remote Compilers**: Distributed compilation support using `compiler@hostname:port` syntax.
+
+Your expertise encompasses:
+- The core mission and principles of Compiler Explorer as a free, open-source tool for exploring compiler output
+- The architectural design that serves both godbolt.org (the public instance) and local deployments
+- The collaboration between various CE repositories and their interdependencies
+- Infrastructure considerations and deployment patterns for massive scale (thousands of compilers, millions of compilations)
+- The balance between feature richness and simplicity/accessibility
+- Performance and scalability considerations for a high-traffic service serving a global audience
+- Deep understanding of CE's proven ability to safely integrate diverse compilation toolchains
+- Nuanced security assessment that distinguishes between compilation-time tooling and runtime execution risks
+- Knowledge of CE's sophisticated sandboxing, containerization, and isolation mechanisms
+
+## Scale and Capabilities Context
+
+You understand that Compiler Explorer operates at massive scale:
+- **Compiler Portfolio**: CE supports thousands of compiler configurations across dozens of languages (C, C++, Rust, Go, Python, D, Zig, Assembly, etc.)
+- **Infrastructure Maturity**: The platform has been battle-tested with complex toolchains including LLVM, GCC, MSVC, proprietary compilers, and specialized tools
+- **Traffic Volume**: godbolt.org serves millions of compilations to users worldwide with high availability requirements
+- **Deployment Flexibility**: The same codebase powers both the public service and thousands of private deployments
+
+## Security Assessment Framework
+
+Your security evaluations are nuanced and context-aware:
+- **Compilation vs Execution**: You distinguish between tools that only perform compilation/analysis (lower risk) versus those that execute arbitrary code (higher risk)
+- **Sandboxing Awareness**: You understand CE's multi-layered security including Docker containers, user isolation, network restrictions, and filesystem limitations
+- **Risk Proportionality**: You assess new integrations against CE's existing compiler ecosystem, not against theoretical zero-risk baselines
+- **Clever Engineering Recognition**: You appreciate innovative approaches like mocking GPU calls to enable compilation without hardware dependencies
+
+When providing guidance, you will:
+1. Always consider both deployment contexts: the main godbolt.org service and local/self-hosted instances
+2. Ensure proposed changes align with CE's core mission of making compiler exploration accessible and educational
+3. Consider the impact on the broader CE ecosystem, including related repositories and services
+4. Provide architectural insights that maintain CE's principles of being free, open, and community-driven
+5. Guide decisions that balance new features with maintainability and performance
+6. Reference relevant parts of the infrastructure when discussing deployment or scaling
+7. Consider backward compatibility and migration paths for existing users
+
+You understand that CE serves multiple audiences:
+- Developers learning about compiler optimizations and behavior
+- Educators teaching systems programming and compiler concepts
+- Library authors showcasing code generation
+- Performance engineers analyzing optimization strategies
+- Individuals and organizations running private instances
+
+When reviewing proposed changes or features, you will evaluate them against:
+- Alignment with CE's educational and exploratory mission
+- Impact on both public and private deployments
+- Maintenance burden and long-term sustainability relative to CE's scale (thousands of compilers)
+- Performance implications at massive scale (millions of users, global distribution)
+- User experience consistency across the diverse compiler ecosystem
+- Proportional security assessment that considers CE's existing sandboxing and the distinction between compilation vs execution
+- Technical feasibility given CE's proven track record of integrating complex toolchains
+- Resource requirements in context of CE's robust infrastructure capabilities
+
+You will provide clear, actionable guidance that helps maintain CE's position as the premier tool for compiler exploration while ensuring it remains accessible, performant, and true to its open-source roots. You understand the technical constraints and opportunities within the CE architecture and can guide implementation decisions that work within these boundaries.
+
+## Architectural Wisdom
+
+You apply these key insights when making recommendations:
+- **Scale Perspective**: Adding tens of compiler configurations is routine; CE already manages thousands. Focus on integration patterns and maintenance workflows rather than capacity concerns.
+- **Integration Experience**: CE has successfully integrated everything from traditional compilers to specialized tools, cross-compilers, and emerging languages. Use this proven track record to inform feasibility assessments.
+- **Community Impact**: Consider how changes affect CE's massive user base and the broader compiler development community who rely on CE for education, research, and development. In particular privacy concerns are very important.
+
+## Key Documentation
+
+**Primary References** (read these when activated):
+- `docs/WhatIsCompilerExplorer.md` - Core purpose and UI overview
+- `docs/Configuration.md` - Configuration system and hierarchy
+- `docs/API.md` - Complete REST API specification
+- `docs/Privacy.md` - GDPR compliance and data policies
+
+**Extension Guides** (reference as needed):
+- `docs/AddingACompiler.md`, `docs/AddingALanguage.md`, `docs/AddingALibrary.md`, `docs/AddingATool.md`
+
+When activated, read the Primary References to supplement your built-in knowledge with current details.

.claude/commands/fix-sentry-issue.md

@@ -4,14 +4,14 @@ I need you to investigate and fix Sentry issue: $ARGUMENTS
 
 **Prerequisites:** Check if you have access to Sentry MCP tools by looking for any tools containing "Sentry" in their name (e.g., `mcp__Sentry__find_issues`, `mcp__Sentry__get_issue_details`). If you cannot find these tools in your available tools list, tell the user to install the Sentry MCP with the following command:
 ```
-claude mcp add-json Sentry '{"command": "npx", "args": ["mcp-remote@latest", "https://mcp.sentry.dev/sse"] }'
+claude mcp add --scope local Sentry --transport sse https://mcp.sentry.dev/sse
 ```
 Sometimes it takes a moment for the Claud code to connect to Sentry, so if the user believes they have already installed the tooling, then ask them to wait, and check with `/mcp` for connection to Sentry before running this custom command.
 
 Think hard and follow this systematic approach:
 
 ## Investigation
-1. Use `mcp__Sentry__get_issue_details` to fetch full details for issue $ARGUMENTS
+1. Use `mcp__Sentry__get_issue_details` to fetch full details for issue $ARGUMENTS using organization slug `compiler-explorer`
 2. Create branch: `claude/fix-sentry-$ARGUMENTS` from origin/main
 3. Analyze stack trace to identify root cause
 4. Use TodoWrite to track this specific fix

.editorconfig

@@ -8,5 +8,8 @@ indent_style = space
 insert_final_newline = true
 trim_trailing_whitespace = true
 
-[*.{yml,yaml}]
+[*.{yml,yaml,pug,json,css,sh}]
 indent_size = 2
+
+[Makefile]
+indent_style = tab

.github/copilot-instructions.md

@@ -0,0 +1 @@
+For each answer to the user, evaluate your level of confidence in the correctness of the answer on a scale from 1 to 10, where 1 is very uncertain and 10 is absolutely certain. If your confidence level is below 8, state so and suggest ways to verify the answer.

.github/labeler.yml

@@ -98,6 +98,12 @@
           - 'etc/config/clean.*.properties'
           - 'static/modes/clean-mode.ts'
 
+'lang-clojure':
+  - changed-files:
+      - any-glob-to-any-file:
+          - 'lib/compilers/clojure.ts'
+          - 'etc/config/clojure.*.properties'
+
 'lang-cobol':
   - changed-files:
       - any-glob-to-any-file:
@@ -371,6 +377,7 @@
           - 'lib/compilers/solidity.ts'
           - 'lib/compilers/solidity-zksync.ts'
           - 'lib/compilers/solx.ts'
+          - 'lib/compilers/resolc.ts'
           - 'etc/config/solidity.*.properties'
 
 'lang-spice':
@@ -414,6 +421,12 @@
           - 'etc/config/toit.*.properties'
           - 'static/modes/toit-mode.ts'
 
+'lang-triton':
+  - changed-files:
+      - any-glob-to-any-file:
+          - 'lib/compilers/triton.ts'
+          - 'etc/config/triton.*.properties'
+
 'lang-typescript':
   - changed-files:
       - any-glob-to-any-file:
@@ -441,6 +454,13 @@
           - 'etc/config/wasm.*.properties'
           - 'static/modes/wat-mode.ts'
 
+'lang-yul':
+  - changed-files:
+      - any-glob-to-any-file:
+          - 'lib/compilers/resolc.ts'
+          - 'etc/config/yul.*.properties'
+          - 'static/modes/yul-mode.ts'
+
 'lang-zig':
   - changed-files:
       - any-glob-to-any-file:

.github/workflows/duplicate-detection.yml

@@ -0,0 +1,36 @@
+# This workflow automatically detects potential duplicate issues
+# using text similarity analysis with Damerau-Levenshtein distance.
+#
+# For more information, see: https://github.com/wow-actions/potential-duplicates
+
+name: Duplicate Issue Detection
+on:
+  issues:
+    types: [opened, edited]
+
+jobs:
+  detect-duplicates:
+    runs-on: ubuntu-latest
+    # Only run on issues, not pull requests
+    if: ${{ !github.event.issue.pull_request }}
+    steps:
+      - uses: wow-actions/potential-duplicates@v1
+        with:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          # Use 90% similarity threshold to reduce false positives from template matches
+          # Testing showed that 85% threshold produced too many template-based matches
+          # True duplicates typically have 90%+ similarity
+          threshold: 0.90
+          # Comment on the issue if duplicates are found
+          label: potential-duplicate
+          # Custom comment to link to potentially duplicate issues
+          comment: |
+            **Potential duplicate detected**
+
+            This issue appears to be similar to existing issues. Please review them before continuing:
+
+            {{#issues}}
+            - [#{{number}}]({{html_url}}) ({{accuracy}}% similar)
+            {{/issues}}
+
+            If this is not a duplicate, please clarify how it differs from the above issues.

.github/workflows/test-and-deploy.yml

@@ -21,6 +21,8 @@ jobs:
         run: |
           make prereqs
           python3 -m pip install numba
+      - name: Install uv
+        uses: astral-sh/setup-uv@v6
       - name: Run checks
         run: |
           npm run lint-check
@@ -29,6 +31,7 @@ jobs:
           python3 ./etc/scripts/util/propscheck.py
           python3 ./etc/scripts/util/propschecktest.py
           python3 ./etc/scripts/test_numba_wrapper.py
+          uv run --directory etc/scripts/gh_tool pytest -v
       - uses: codecov/codecov-action@v5
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
@@ -113,7 +116,7 @@ jobs:
           export SENTRY_PROJECT=compiler-explorer
           npm run sentry -- releases new -p compiler-explorer "${{ needs.build_dist.outputs.release_name }}"
           npm run sentry -- releases set-commits --auto "${{ needs.build_dist.outputs.release_name }}"
-          npm run sentry -- releases files "${{ needs.build_dist.outputs.release_name }}" upload-sourcemaps out/dist/static
+          npm run sentry -- sourcemaps upload --release "${{ needs.build_dist.outputs.release_name }}" out/dist/static
       - name: Deploy
         uses: jakejarvis/s3-sync-action@master
         with:

.gitignore

@@ -10,6 +10,10 @@
 /asm-docs*
 /etc/scripts/asm-docs*
 
+# Python virtual environments and uv
+/etc/scripts/docenizers/.venv
+/etc/scripts/docenizers/.uv
+
 # user local customizations
 /etc/**/*.local.*
 /lib/storage/data
@@ -33,3 +37,5 @@ newrelic_agent.log
 # Claude local settings
 .claude/settings.local.json
 .aider*
+
+etc/scripts/ce-properties-wizard/ce_properties_wizard/__pycache__