agentleFS
jsuchetjsuchet/jonathan-skills-lab/skills/brain-save/SKILL.md

brain-save

Skill1 starsChanged 5 months agoMITView on GitHub

Save an insight or knowledge from the current conversation into the brain. Lighter version of /brain-add designed for in-context capture.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 12 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    By Claude: When the CARL brain-capture rule fires and user approves a suggested save

    line 17
  2. Distinctive:

    If argument provided: use it as the core content/description

    line 24
  3. Distinctive:

    If no argument: review the recent conversation context and identify the most brain-worthy insight, learning, or reference discussed

    line 25
  4. Distinctive:

    If invoked after a brain-capture suggestion, save the specific item suggested

    line 26
  5. Distinctive:

    Add 3-5 inline [[wiki-links]] for key concepts

    line 40
  6. Distinctive:

    Keep it focused — one insight per note.

    line 41
  7. Distinctive:

    Check for duplicates first — if similar note exists, suggest updating it instead

    line 45
  8. Very specific:

    Add to related: if strong matches found

    line 49
  9. Very specific:

    Add new entry in the appropriate category section: - [[note-name]](category/subcategory/note-name.md) — One-line summary.

    line 53
  10. Distinctive:

    Update the note count in the header and the section header count

    line 54
  11. Distinctive:

    Update "Last updated: YYYY-MM-DD"

    line 55
  12. Distinctive:

    Use Bash to append directly without reading:

    line 58
  13. Distinctive:

    Keep notes atomic — one concept per note.

    line 74
  14. Distinctive:

    Write the note as standalone knowledge, not as "during our conversation we discussed..." — strip conversation context.

    line 75

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/brain-save [description]

Capture knowledge from the current conversation and file it in the brain.

When this is invoked

  • By user: /brain-save or /brain-save "description of what to save"
  • By Claude: When the CARL brain-capture rule fires and user approves a suggested save

Steps

  1. Read brain-config.md for taxonomy and filing rules.
  2. Identify what to save:
    • If argument provided: use it as the core content/description
    • If no argument: review the recent conversation context and identify the most brain-worthy insight, learning, or reference discussed
    • If invoked after a brain-capture suggestion, save the specific item suggested
  3. Determine note type and category:
    • Insight/opinion → note
    • Factual information → reference
    • How-to discovered in conversation → guide or process
    • Idea sparked → idea
    • Cross-referenced answer synthesizing multiple brain notes → synthesis (use source: "brain-query", list consulted notes in related:)
    • Match against filing signals for category/subcategory
  4. Create a concise, well-structured note:
    • YAML frontmatter with source: "conversation" and today's date
    • Body: distill the insight into clear, standalone prose (not conversation transcript)
    • The note should make sense to someone reading it weeks later without conversation context
    • Add 3-5 inline [[wiki-links]] for key concepts
    • Keep it focused — one insight per note. If multiple things to save, create multiple notes.
  5. File the note:
    • Path: brain/{category}/{subcategory}/{kebab-case-title}.md
    • Check for duplicates first — if similar note exists, suggest updating it instead
  6. Quick connection check:
    • Grep for shared tags/topics in existing notes
    • Add to related: if strong matches found
  7. Update index.md:
    • Read brain/index.md
    • Add new entry in the appropriate category section: - [[note-name]](category/subcategory/note-name.md) — One-line summary. Tags: tag1, tag2
    • Update the note count in the header and the section header count
    • Update "Last updated: YYYY-MM-DD"
  8. Append to log.md (do NOT read the file — append only):
    • Use Bash to append directly without reading:
      echo '## [YYYY-MM-DD] ingest | {note-title}
      Filed to: {category}/{subcategory}. Source: conversation. {X} wiki-links created.
      ' >> brain/meta/log.md
  9. Confirm:
    🧠 Saved to brain/{category}/{subcategory}/{filename}.md
    Tags: #tag1 #tag2

Important

  • This skill can be invoked by Claude (disable-model-invocation: false) when the brain-capture CARL rule suggests a save and the user confirms.
  • Keep notes atomic — one concept per note. Suggest multiple /brain-save calls if there are several things to capture.
  • Write the note as standalone knowledge, not as "during our conversation we discussed..." — strip conversation context.
  • Be brief. A brain-save note should be 3-15 lines of body content typically.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.