Civic Transparency Software Development Kit (SDK)

Synthetic data generation toolkit for civic transparency research and testing.

Installation

pip install civic-transparency-sdk

What This Package Provides

• Synthetic Data Generation: Create realistic transparency data for testing and research without requiring real user data. Generate controlled datasets with reproducible seeds for studying information dynamics.
• Internal Data Structures: Simulation-specific types (WindowAgg, ContentHash, TopHash) for generating and manipulating synthetic transparency data.
• Database Integration: Convert generated data to DuckDB/SQL databases for analysis with ready-to-use schemas and indexing patterns.
• CLI Tools: Command-line utilities for generating worlds, converting formats, and managing synthetic datasets.

Note: This SDK generates synthetic data for research/testing. For implementing the PTag API specification, see civic-transparency-ptag-types, which provides the official API response types.

Quick Start

Generate synthetic data:

# Activate environment
source .venv/bin/activate  # Linux/Mac
# or
.venv\Scripts\activate     # Windows

# Generate baseline world
ct-sdk generate --world A --topic-id aa55ee77 --out world_A.jsonl

# Convert to database
ct-sdk convert --jsonl world_A.jsonl --duck world_A.duckdb --schema schema/schema.sql

The generated DuckDB files are ready for analysis with any SQL-compatible tools or custom analysis scripts.

Use Cases

• Academic Research: Generate controlled datasets with known parameters for studying information dynamics, coordination patterns, and transparency system behaviors.
• Algorithm Development: Build and test transparency tools using synthetic data that mimics real-world patterns without privacy concerns.
• Testing & Validation: Create reproducible test datasets for developing transparency systems without requiring real user data.
• Education: Provide realistic datasets for teaching transparency concepts, data analysis, and system design.

---

Reproducibility

All generation is deterministic:

• Seed-based randomization: Same seed produces identical datasets
• Version tracking: Metadata includes package versions
• Parameter logging: All generation settings preserved in output
• Schema versioning: Database structures fully documented

Example seeds:

• World A (baseline): 4242
• World B (influenced): 8484

---

Package Structure

ci.transparency.sdk/
├── cli/            # Command-line interface (ct-sdk)
├── digests.py      # Content fingerprinting (SimHash64, MinHashSig)
├── hash_core.py    # Content identification (HashId, ContentHash, TopHash)
├── ids.py          # ID management (WorldId, TopicId)
├── io_schema.py    # JSON serialization utilities
└── window_agg.py   # Window aggregation structure (WindowAgg)

---

Related Projects

• Civic Transparency PTag Spec - Official API specification
• Civic Transparency PTag Types - Python types for PTag API responses (use this for API implementation)
• Civic Transparency Verify - Statistical verification tools (private)

---

Security Model

This package provides synthetic data generation for research and testing. It does not include:

• Detection algorithms or thresholds
• Verification workflows or assessment criteria
• Operational patterns or alerting rules

These are maintained separately to prevent adversarial reverse-engineering while enabling legitimate transparency research.

---

Documentation

Full documentation at: civic-interconnect.github.io/civic-transparency-py-sdk/

• Usage Guide - Getting started and common workflows
• CLI Reference - Command-line interface details
• SDK Reference - Core Python APIs
• Schema Reference - Database schemas and integration

---

Development

See CONTRIBUTING.md for development setup and guidelines.

Versioning

This specification follows semantic versioning. See CHANGELOG.md for version history.

License

MIT © Civic Interconnect