agent-research-aggregator

What problem does it solve?

This skill automatically finds and consolidates fragmented experiment artifacts produced by AI coding agents into the structured idea.md and experimental_log.md files PaperOrchestra requires, saving researchers from manual log sifting and error-prone transcription.

Core Features & Use Cases

• Discovery: Read-only scan of common agent caches (.claude, .cursor, .antigravity, .openclaw) and project roots to locate experiment files and metrics.
• Extraction & Synthesis: Batch LLM-assisted extraction of structured experiment records, consolidation into a single synthesis JSON, and deterministic formatting into PaperOrchestra inputs.
• Validation & Audit: Schema validation, truncation safeguards, confidence tagging, and an optional aggregation report highlighting data quality, file counts, and warnings.
• Use Case: Turn scattered experiment notes, chat histories, and metric files from multiple coding-agent workspaces into a reviewable idea.md and experimental_log.md so PaperOrchestra can generate paper content.

Quick Start

Ask the agent: Aggregate my agent logs from ~/projects into PaperOrchestra inputs and generate workspace/inputs/idea.md and workspace/inputs/experimental_log.md.