Agent-Skills.md

Agent Skills: Second Brain Skill

Personal intelligence system for capturing thoughts, managing knowledge, and surfacing insights. Use when user wants to capture an idea, task, or note during conversation; query their knowledge base; check their inbox; review digests; or update task status. Triggers include "remember this," "add a task," "what did I say about," "show my inbox," or "mark complete."

UncategorizedID: leegonzales/aiskills/second-brain

Repository

leegonzales/aiskills
leegonzalesLicense: MIT
255

Install this agent skill to your local

pnpm dlx add-skill https://github.com/leegonzales/AISkills/tree/HEAD/SecondBrain/second-brain

Skill Files

Browse the full folder contents for second-brain.

Download Skill

Loading file tree…

SecondBrain/second-brain/SKILL.md

Skill Metadata

Name
second-brain
Description
Personal intelligence system for capturing thoughts, managing knowledge, and surfacing insights. Use when user wants to capture an idea, task, or note during conversation; query their knowledge base; check their inbox; review digests; or update task status. Triggers include "remember this," "add a task," "what did I say about," "show my inbox," or "mark complete."

Second Brain Skill

Conversational interface to the Second Brain personal knowledge management system. Capture thoughts naturally during Claude Code sessions, query your knowledge graph, and manage your inbox.

Core philosophy: Capture at the point of thinking, not after. The system remembers so you don't have to.

When to Use

Recognize and route these intents:

| Intent | Trigger phrases | Command | |--------|-----------------|---------| | Capture | "remember this", "add a task", "note:", "capture this idea" — also implicit: "I should...", "don't forget..." | sb capture | | Query | "what did I say about...", "what supports goal X", "show tasks related to..." | sb list / query | | Inbox | "show my inbox", "what's waiting for review", "how many pending" | sb inbox | | Digest | "what should I focus on today", "today's digest", "what's overdue" | sb digest | | Action | "mark done", "complete task X", "archive...", "set priority to..." | sb show / update | | Transcript | user pastes a meeting transcript to extract structured content | see ref |

CLI Commands (load-bearing)

This skill wraps the sb CLI:

sb capture "content"      # Capture a thought (--source cli|slack|email|calendar|file)
sb inbox                  # List pending captures (--status, --limit)
sb process                # Classify pending captures (--id, --dry-run)
sb digest                 # Generate daily digest
sb list [type]            # List nodes (--status, --domain, --limit)
sb show <id>              # Show node details (by ID or unique title)
sb status                 # System health check
sb init                   # Initialize DB and directories (--force)

Full flag tables, output examples, env vars, and exit codes: see references/cli-reference.md.

Quick Action Shortcuts

/sb capture "thought or idea"     # Capture immediately
/sb inbox                         # Show pending items
/sb digest                        # Today's actionable summary
/sb query "search term"           # Search knowledge base
/sb done <id>                     # Mark task complete

Classification

Captures are AI-classified into node types: task, idea, reference, meeting, goal, project, value, person. Confidence threshold gates the inbox:

  • High confidence (>=0.6): auto-classified
  • Low confidence (<0.6): sent to needs_review

Node-type definitions, the classification decision tree, priority levels (0-4), and domain heuristics: see references/node-types.md.

Capture Workflow

When something capture-worthy comes up in conversation:

  1. Recognize intent — direct ("Remember this...") or implicit ("I should...").
  2. Capture with context — include relevant conversation context, tag --source cli, add any mentioned relationships.
  3. Confirm — brief confirmation with ID; flag if needs_review.

Be selective — capture on expressed intent or importance, not every utterance. Detailed workflow, context-awareness rules, anti-patterns, transcript-processing extraction patterns, and worked examples: see references/workflows-and-examples.md.

Graph Model (summary)

Nodes are typed (8 node types above) and connected by typed edges:

| Edge | Meaning | |------|---------| | supports | provides evidence for (project -> goal) | | blocks | prevents progress (task -> task) | | contains | hierarchical parent (project -> task) | | derived_from | extracted from (goal -> value) | | assigned_to | assigned to person (task -> person) | | mentioned_in | referenced in context (person -> meeting) | | related_to | general relationship (idea -> reference) | | child_of | subtask/child (task -> task) |

Implementation

  • Database: local SQLite at ~/.local/share/secondbrain/secondbrain.db
  • Vault: Obsidian markdown output (wikilinks, daily notes include digest)
  • Config: ~/.config/secondbrain/config.yml
# ~/.config/secondbrain/config.yml
node_id: "home"
vault_path: "/path/to/vault"
classification:
  model: "claude-sonnet-latest"  # use the current Sonnet ID; don't pin a dated value
  confidence_threshold: 0.6

Reference Files

  • references/cli-reference.md — full sb command/flag tables, output examples, environment variables, file locations, exit codes.
  • references/node-types.md — node-type definitions, classification decision tree, priority levels, domain classification heuristics.
  • references/workflows-and-examples.md — workflow integration, context awareness, meeting-transcript extraction patterns + config, worked example interactions, anti-patterns, integration points, success metrics.