CCL Prompts for AI Agents
CCL Prompts for AI Agents
Section titled “CCL Prompts for AI Agents”Copy-paste these prompts to guide AI agents building CCL implementations.
How to Use
Section titled “How to Use”- Choose a prompt that matches your goal
- Replace
[language]with your target language (Rust, Python, etc.) - Paste into your AI agent (Claude, ChatGPT, Copilot, etc.)
- The AI will read the docs and follow the specification
Full Implementation
Section titled “Full Implementation”Basic Implementation
Section titled “Basic Implementation”Build a CCL parser in [language].
Read https://ccl.tylerbutler.com/ai-implementation-guide/ for the completespecification. Start with parse and build_hierarchy, then add typed access.
Use tests from https://github.com/tylerbutler/ccl-test-data to validate.With Specific Behaviors
Section titled “With Specific Behaviors”Build a CCL parser in [language] with these behaviors:- crlf_normalize_to_lf (normalize line endings)- boolean_strict (only "true" and "false")- list_coercion_disabled (require explicit list syntax)
Read https://ccl.tylerbutler.com/ai-implementation-guide/ for the specification.See https://ccl.tylerbutler.com/behavior-reference/ for behavior details.Minimal Core Only
Section titled “Minimal Core Only”Build a minimal CCL parser in [language] with only the required functions:- parse (text to entries)- build_hierarchy (entries to nested object)
Skip typed access and formatting for now.
Read https://ccl.tylerbutler.com/ai-implementation-guide/#core-functions-requiredFeature-Specific
Section titled “Feature-Specific”Add Typed Access
Section titled “Add Typed Access”I have a CCL parser with parse and build_hierarchy working.
Add typed access functions:- get_string- get_int- get_bool- get_float- get_list
Follow https://ccl.tylerbutler.com/ai-implementation-guide/#typed-access-functions-optionalUse https://ccl.tylerbutler.com/library-features/ for implementation details.Add Single Function
Section titled “Add Single Function”Add the get_list function to my CCL implementation following the spec athttps://ccl.tylerbutler.com/ai-implementation-guide/#get_list
Use list_coercion_disabled behavior (error if value is not actually a list).Add Formatting
Section titled “Add Formatting”Add formatting functions to my CCL implementation:- print (structure-preserving round-trip)- canonical_format (normalized output)
Follow https://ccl.tylerbutler.com/library-features/#formatting-functionsAdd Entry Processing
Section titled “Add Entry Processing”Add entry processing functions to my CCL implementation:- filter (remove entries by predicate, e.g., comments)- compose (concatenate entry lists)
Follow https://ccl.tylerbutler.com/library-features/Test Integration
Section titled “Test Integration”Set Up Test Suite
Section titled “Set Up Test Suite”Set up test integration for my CCL implementation using the test suite athttps://github.com/tylerbutler/ccl-test-data
The tests are in generated_tests/*.json format. Filter tests to only runthose for functions I've implemented: parse, build_hierarchy.
See https://ccl.tylerbutler.com/test-suite-guide/ for test format details.Build Test Runner
Section titled “Build Test Runner”Build a test runner for my [language] CCL implementation that:1. Loads JSON tests from https://github.com/tylerbutler/ccl-test-data/generated_tests/2. Filters by my implemented functions: [list your functions]3. Runs each test and reports pass/fail4. Handles the "expected" format with "count" fields
See https://ccl.tylerbutler.com/test-suite-guide/ for the test format.Run Specific Test Categories
Section titled “Run Specific Test Categories”Run only the parse function tests from the CCL test suite.
Filter tests where validation === "parse" from the generated_tests/ directory.Report which tests pass and fail.Debugging & Maintenance
Section titled “Debugging & Maintenance”Fix Failing Tests
Section titled “Fix Failing Tests”My CCL parser is failing these tests: [paste test names or errors]
Review my implementation against the algorithm athttps://ccl.tylerbutler.com/parsing-algorithm/ and identify issues.
Common pitfalls are listed athttps://ccl.tylerbutler.com/ai-implementation-guide/#common-pitfallsReview Implementation
Section titled “Review Implementation”Review my CCL implementation for correctness.
Check against:- Algorithm: https://ccl.tylerbutler.com/parsing-algorithm/- Common pitfalls: https://ccl.tylerbutler.com/ai-implementation-guide/#common-pitfalls- Behaviors: https://ccl.tylerbutler.com/behavior-reference/
Identify any deviations from the specification.Add Missing Error Handling
Section titled “Add Missing Error Handling”Review my CCL implementation's error handling.
The spec at https://ccl.tylerbutler.com/ai-implementation-guide/ describesexpected error cases. Ensure my implementation handles:- Missing keys (path not found)- Type mismatches (e.g., get_int on non-integer)- Malformed input (lines without =)Quick Prompts
Section titled “Quick Prompts”One-Liner: Start Fresh
Section titled “One-Liner: Start Fresh”Implement CCL in [language] following https://ccl.tylerbutler.com/ai-implementation-guide/One-Liner: Quick Reference
Section titled “One-Liner: Quick Reference”Read https://ccl.tylerbutler.com/ai-quickstart/ for CCL terminology and concepts.One-Liner: Check My Work
Section titled “One-Liner: Check My Work”Compare my CCL parser to the spec at https://ccl.tylerbutler.com/parsing-algorithm/Prompt Tips
Section titled “Prompt Tips”Be specific about functions: Instead of “add more features,” say “add get_string and get_int.”
Specify behaviors: If you care about boolean parsing or list coercion, say so explicitly.
Reference specific URLs: Point to exact sections like #get_list for focused work.
Include your constraints: Language version, dependencies, style preferences.
Mention the test suite: Always reference https://github.com/tylerbutler/ccl-test-data for validation.