Claude Cpp Agent Pipeline
Description
“Built a Claude SDK agentic pipeline to automate C++ normalization manager generation for financial exchanges.” What does this mean? Financial exchanges have complicated specifications describing how market/exchange data should be processed.
Installation
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open the source below and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
README
Claude SDK Agentic Pipeline for C++ Code Generation
An educational/portfolio implementation of an agentic pipeline that converts structured exchange specifications into deterministic C++ normalization-manager code.
**Portfolio note:** This repository uses synthetic, exchange-like specifications. It does not contain proprietary code, schemas, data, or internal implementation from any employer.
Architecture
flowchart TD
A[Exchange Specification] --> B[Parser Agent]
B --> C[Mapping Agent]
C --> D[Ambiguity Resolution]
D --> E[Jinja2 C++ Generator]
E --> F[Pre-Compilation Validator]
F -->|Valid| G[C++ Compiler]
F -->|Invalid| H[Repair Agent]
G -->|Pass| I[Regression Oracle]
G -->|Fail| H
I -->|Pass| J[Accepted Artifact]
I -->|Fail| H
H --> K[Patch + Retry]
K --> E
What this project demonstrates
- Claude SDK / Anthropic Messages API integration
- Agentic decomposition of a code-generation problem
- Structured JSON outputs from LLM agents
- Protocol-field mapping and ambiguity resolution
- Deterministic C++ generation with Jinja2
- Structural/type contract validation before compilation
- Automated compile/test feedback
- Regression-oracle based acceptance/rejection
- Bounded autonomous repair loop
- Offline mode for reproducible demos and tests
Repository structure
claude-cpp-agent-pipeline/
├── app.py
├── requirements.txt
├── .env.example
├── .gitignore
├── README.md
├── config/
│ └── pipeline.yaml
├── specs/
│ └── sample_exchange.yaml
├── templates/
│ └── normalization_manager.cpp.j2
├── src/
│ ├── __init__.py
│ ├── agents.py
│ ├── generator.py
│ ├── validator.py
│ ├── oracle.py
│ └── pipeline.py
└── tests/
├── test_generator.py
└── test_validator.py
Setup
Python 3.10+ is recommended.
git clone https://github.com//claude-cpp-agent-pipeline.git
cd claude-cpp-agent-pipeline
python -m venv .venv
# Windows
.venv\Scripts
Related Skills
mcp-server-postgres
Read-only PostgreSQL database access.
Data mcp-server-sqlite
SQLite database interaction and querying.
Data mcp-server-google-maps
Google Maps integration for location data.
Data Bitbucket Data Center
---
Data Private Gpt
Complete API layer for private AI applications on local models: RAG, skills, tools, MCP, text-to-sql, and more
Data Csv Data Summarizer
Automatically analyze CSV files and generate comprehensive insights with visualizations
Data