Save thoughts and content into a GBrain knowledge base via one CLI command.
npx skills add https://github.com/AMC-JTC/gbrain-1 --skill capture-amc-jtc
Please help me install this Agent Skill. Skill: capture Source: https://github.com/AMC-JTC/gbrain-1/tree/main/plugin/skills/capture Command: npx skills add https://github.com/AMC-JTC/gbrain-1 --skill capture-amc-jtc
gbrain capture command that writes content to both the brain database and a markdown file on disk in one step. ## Core Features & Use Cases - Single ingestion entrypoint: Accepts content as an inline argument, a file via --file PATH, or piped input via --stdin, then writes it to the brain DB and disk simultaneously. - Automatic slugging and frontmatter: Generates a stable inbox/YYYY-MM-DD-<hash8> slug, wraps plain prose in sensible frontmatter (type, title, captured_via, captured_at), and deduplicates re-captures via content hashing. - Script-friendly output: Supports --quiet for slug-only output and --json for structured output, making it usable in shell pipelines and agent workflows. - Use Case: After a meeting, you paste a summary and say "save this to my brain" — the Skill captures it as a queryable page under inbox/ with a printed receipt showing slug, hash, and file path. ## Quick Start Ask the assistant to capture a thought by saying "capture this: <your thought>" and it will run gbrain capture to save it into your brain inbox.High-intent search queries and answers about installing and using this skill.
Run `gbrain capture "your thought"` to save content in one command. It writes the page to both the brain database and a markdown file under your sync repo path, making it immediately queryable via gbrain query or search.
gbrain capture is the human-facing wrapper that handles default slug generation, content-type heuristics, frontmatter stamping, and a receipt output. gbrain put is the lower-level per-page primitive without those conveniences, so capture is preferred for saving single thoughts.
Yes, use `gbrain capture --file PATH` for an existing markdown file or pipe content with `echo "text" | gbrain capture --stdin`. These options are also safer than inline arguments for sensitive content, since inline args land in shell history.
Yes, identical content produces the same `inbox/YYYY-MM-DD-<hash8>` slug, and the daemon's 24-hour content-hash dedup catches re-captures. This makes the operation idempotent for repeated saves of the same text.
Avoid capture for bulk importing many files, ingesting articles with author metadata, or processing meeting transcripts with attendee enrichment. Use gbrain sync, the idea-ingest skill, or the meeting-ingestion skill for those specialized cases.
PieroSierraIngest markdown files from raw/ into structured wiki articles and update the index.
AbilityaiDiscover hidden connections and patterns between knowledge base notes using semantic search and network analysis.
garrytanMigrates a GBrain knowledge base between storage formats or versions.
ngochuy13Ingest meeting transcripts into a knowledge graph with enriched attendees and entity links.