research-assistant
Read Bear notes tagged "待整理", extract topic keywords, search for relevant GIFs via gifgrep, insert them into the note, and remove the tag. Use when the user wants to auto-illustrate or finalize research notes from Bear.
Research Assistant
Automate the workflow of enriching Bear research notes with topical GIFs.
Prerequisites
- Bear app running with a valid API token (
~/.config/grizzly/token) grizzlyCLI installed- gifgrep skill available (for GIF search)
Workflow
-
Fetch notes — List all notes with the
待整理tag:bash grizzly open-tag --name "待整理" --enable-callback --json --token-file ~/.config/grizzly/token -
For each note: a. Read the note content via
grizzly open-note --id <ID> --enable-callback --json. b. Extract 2–3 topic keywords from the title and key-findings sections. c. Search for a relevant GIF using gifgrep (orweb_search+web_fetchfor a GIF URL) with those keywords. d. Append the GIF as markdown to the note's "Supporting Media" section:bash echo '' | grizzly add-text --id <ID> --mode append --token-file ~/.config/grizzly/tokene. Remove the
待整理tag by replacing tags (exclude待整理, keep all others). -
Report — Summarize which notes were processed and how many GIFs were inserted.
Notes
- If a note already has content in "Supporting Media", insert the GIF on a new line below existing media.
- If no relevant GIF is found, skip insertion and note it in the report.
- Prefer GIFs that visually represent the research topic (e.g., data visualization, concept animation).
- The
待整理tag signals "needs processing"; removing it marks the note as finalized.
Related skills
- Downloads
- Security audit
- Pass
- Last updated
- 6mo ago
- Current version
- v0.1.0
- License
- MIT-0
