brain-load

name

brain-load

description

Load project context from the Claude Brain Logseq graph into the current session. Triggers: "load brain", "load <project>", "resume <project>", "continue work on <project>", "what do we know about <topic>". Don't fire for write operations (use brain-save), generic questions about Logseq itself, or "open <file>" / "switch to <branch>" requests that mean opening files or switching git branches rather than loading project memory.

Brain Load

Read project context from the Claude Brain Logseq graph and present it to Claude so work can continue seamlessly across sessions and devices.

Prerequisites

Resolve the graph path per skills/_shared/path-resolution.md.

Loading a Specific Project

When the user says "load " or similar:

1. Find the project page using the algorithm in references/matching.md.

2. Read the matched project page using the section-targeted-read pattern in skills/_shared/section-locator.md. In brief mode, read only Properties + Overview + Current Plan + last 3 Session Log entries (target ≤80 total lines). In full mode, read all sections but still section-by-section, never as a single big Read.

3. Read related context (full mode only). Follow [[link]] references inside the project page if they're critical (e.g., a linked design doc). Skip in brief mode.

4. Read today's journal (journals/yyyy_MM_dd.md) if it exists — captures session notes from earlier today.

5. Check active Jira tasks. If the project's Current Plan section references task IDs (e.g., PROJ-1234), and the task folder is accessible, optionally read plan.md. Full mode or explicit user ask only — in brief mode, just mention task IDs and status. When a referenced task has its own page (pages/Tasks___<ID>.md), check its page-top status::: in brief mode skip done tasks entirely; in full mode mention them as "done" without loading their content.

6. Apply staleness rules. Use skills/_shared/staleness.md against the project's last-updated:: and status::.

7. Surface session continuity hints. From the most recent Session Log entry: what was last worked on, any open-questions::, any blockers. Frame as: "Picking up where you left off — last session ([date]): [summary]. Open questions: [list]."

8. Present a context summary conversationally: project name + status + staleness, current plan/task, continuity hint, key recent decisions (full mode only), suggested next actions.

9. Write a journey-log entry per skills/_shared/journey-log.md with activity line: loaded [[Projects/<MatchedName>]] (brief) or loaded [[Projects/<MatchedName>]] (full).

Loading General Context

When the user says "load brain" without a project:

1. Read pages/Index.md for the list of all projects.
2. Read pages/Meta.md for user preferences and conventions.
3. Present a high-level overview: active projects, cross-project decisions, user preferences.
4. Write a journey-log entry: loaded brain overview.

Searching Across the Brain

When the user asks "what do we know about X" or similar, follow the algorithm in references/search.md. After presenting findings, write a journey-log entry: searched "X" · N hits. Label hits from done-task pages and session-archive pages as such in the results, so cold context is recognizable.

Load Modes: Brief vs Full

Brief mode (default — "load "):

1. Properties (type, status, last-updated)
2. Overview (first 5 bullets only)
3. Current Plan (full)
4. Session Log (last 3 entries only)
5. Skip Implementation, Decisions, linked context, done task pages, and session-archive pages (type:: session-archive)

Full mode ("load full", "load everything about "):

1. All properties
2. Full Overview
3. Full Current Plan
4. Full Implementation
5. All Decisions
6. Last 10 Session Log entries
7. Related context pages (follow [[links]])
8. Today's journal entry

In brief mode, mention that more is available: "Loaded brief context. Say 'load full' for decisions, implementation details, and full history."

Important Notes

• Be selective about reads. Token budget matters — use the section-targeted-read pattern in step 2 of "Loading a Specific Project" rather than reading whole pages.
• After loading, confirm what's available and suggest next actions.
• All graph content is Logseq outliner format. See CLAUDE.md for invariants.