nature-experiment-log

nature-experiment-log is a research-workflow skill (documented in Chinese and English) that standardizes lab experiment logging. It accepts raw material such as images, voice transcripts, and text and produces structured Markdown experiment logs with YAML frontmatter into an Obsidian vault or plain Markdown workflow. It solves the problem of scattered, inconsistent lab records by turning a day's observations into traceable, searchable logs that pair with Obsidian's Dataview plugin for dynamic queries and dashboards.

The skill defines a full processing pipeline: it reads and analyzes provided materials to extract structured information, generates experiment and sample-batch IDs following documented naming rules (a system code, equipment code, date, and sequence number), writes a standard log into a wiki/experiment-log directory tree, archives raw materials into a dated raw/experiments folder, appends anomaly records and index entries, and reports where files were written. It notably refuses to invent temperature, duration, recipe, device model, or results, preserving unconfirmed fields as items needing author input and proactively asking about ambiguous data rather than guessing. It ships templates (experiment index, anomaly log, equipment tracking) and three complete example logs covering corrosion, electrochemical, and thermal-stability experiments. An optional path integrates with a separate feishu-cli-integration skill to pull messages and download image resources from a Feishu research group using a lark-cli command-line tool, requiring a configured bot with im:message and im:resource permissions.

Target users are experimental researchers and lab teams (materials, electrochemistry, thermal) who keep notebooks in Obsidian and want consistent, FAIR-friendly, version-controllable records. External write paths depend on the user's own local CLI configuration and directory permissions.