From ai-brain
Weekly synthesis of your brain thoughts, workflow insights, and goals. Surfaces gaps, open loops, and recommendations. Every claim is cited back to its source.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ai-brain:weekly-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
A weekly synthesis that cross-references your brain thoughts, workflow insights, and goals to surface what you'd miss looking at any one source alone.
A weekly synthesis that cross-references your brain thoughts, workflow insights, and goals to surface what you'd miss looking at any one source alone.
Every claim in the output must be grounded in a source — cite thoughts as thought:<id>, insights as insight:<id>, lists as list:<id>.
Compute the start of the week as an epoch-ms timestamp.
Date.now() - 7 * 24 * 60 * 60 * 1000 (rounded down to the start of the day).Call this weekStartMs.
Call mcp__ai-brain__timeline_thoughts with:
aroundMs: weekStartMsbefore: 0after: 50This returns up to 50 compact index rows in chronological order for thoughts captured since the start of the week.
If the array is empty, tell the user: "Your brain has no thoughts captured this week. Try capturing some thoughts, or run /brain-init if your brain is empty." Then stop.
Workflow insights:
Call mcp__ai-brain__get_insights with status: "new", then again with status: "noted". If the tool is unavailable (server doesn't expose insights for this user), note this and skip insight-dependent sections.
Goals and priorities:
Call mcp__ai-brain__get_lists with pinned: true to get the user's stated goals and priorities.
Open items:
Call mcp__ai-brain__get_open_items to get unfinished tracked items across all lists.
From the week's timeline (Step 2), select up to 10 thoughts that look most substantive (by summary + snippet + topics) — the ones you'll want to cite in the synthesis.
Call mcp__ai-brain__get_thoughts with ids: [<up to 10 ids>]. Only these hydrated results can be quoted or paraphrased in the synthesis. The rest of the timeline is still referenceable by thought:<id> + summary.
Generate a report with these 5 sections. Each section should be concise — the entire review should be scannable in 2 minutes. Every factual claim must cite its source.
This Week in 30 Seconds
2-3 sentence headline summary. What was the dominant theme? What stands out?
Cite 2-3 anchor thoughts: thought:<id>.
Attention vs. Intention
Compare workflow insights (what you actually did) against pinned goals (what you intended to do).
Flag:
list:<id>.insight:<id>.insight:<id>.If no workflow insights are available, display instead:
"Install the
radarplugin (/plugin install radar@flippyhead/radar) for time allocation analysis."
Knowledge Captured
Review the themes from saved thoughts this week. Cite each theme to an anchor thought.
Highlight:
insight:<id>.thought:<id> examples.Open Loops
Aggregate unfinished threads from all sources:
list:<id> with item counts.insight:<id>.thought:<id> where the decision was raised.Next Week
2-3 specific, actionable recommendations based on the above. Be forward-looking, not retrospective. Reference specific projects, people, or decisions — with citations — when possible.
After presenting the review, ask: "Want me to save a summary of this review to your brain? This helps track trends across weeks."
If yes, save a condensed version via mcp__ai-brain__capture_thought with format:
"Weekly review (week of [ISO date]): [2-3 sentence summary of key themes, attention vs. intention highlights, and top recommendation]. Grounded in: thought:, thought:, insight:."
Return the new thought:<id> to the user so they can find this review later via /brain-thread or /brain-context.
npx claudepluginhub flippyhead/ai-brain --plugin ai-brainProvides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.