You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Description:
Integrate the Ecosystem Engine's direct data access capabilities with the insights from the GraphRAG engine. This will allow AI agents to cross-reference on-chain transaction data and patterns (from GraphRAG) with live blockchain events, oracle data, and governance proposal information (from Ecosystem Engine) to perform more informed and verifiable operations.
Acceptance Criteria:
AI agents can seamlessly utilize data from both direct ecosystem connectors (e.g., FTSO, project APIs) and the GraphRAG engine (for historical transaction analysis) in a combined manner.
Queries can be constructed that correlate different data types, for example, linking a governance proposal's discussion (from VectorRAG on forums) with subsequent on-chain voting patterns (from GraphRAG or direct node access).
flare_ai_kit/ecosystem/tooling/goldsky.py (New file to be created)
Tasks / Implementation Steps:
Create Key Files
Integrate indexing support for a service like Goldsky Mirror (reach out to the Flare team for Goldsky API keys) for data replication
Find the best approach for integration - either with Goldsky Flow, interactive CLI or definition file. Note: For an SDK the definition file is perhaps the most appropriate
Create a data pipeline injecting chain-level datasets (blocks, logs, traces etc)
Expose the injected data via a single Python class (class Goldsky)
