skills/suggest-tasks/SKILL.md
Scan all projects, OKRs, open PRs, Notion board, and calendar to propose prioritized agent tasks. Creates 'Suggested' cards in Notion Control Center with S3 analysis and agent assignment (Viktor/Luna/Marco/Kai/Maya). Part of Maya's operational loop.
npx skillsauth add razbakov/skills skills/suggest-tasksInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
⚠️ Migration in progress (razbakov/ikigai#73): Control Center has moved from Notion to GitHub Issues + Project v2. When this skill references creating/reading/updating Notion cards or pages, translate to GitHub equivalents:
- Create card →
gh issue create --repo <project-repo> --title "<title>" --label agent:<name> --body "<S3 body>"thengh project item-add 5 --owner razbakov --url <issue-url>- Read cards →
gh issue list --repo <repo> --label agent:<name> --state open(or across repos viagh search issues "org:razbakov label:agent:<name> state:open")- Update card status → move on board:
gh project item-editwith the Status field, or close viagh issue close- Board columns: Inbox → To do → In progress → To review → Done
- Do not call any
notion-*MCP tools — the Notion MCP is disabled.
Reads the full project portfolio state and proposes what AI agents should work on next. Creates "Suggested" cards in the Notion Control Center for Alex to approve before dispatch.
User says /suggest-tasks, "suggest tasks", "what should agents work on", or auto-invoked as daily-review step 6.5.
~/Projects/ikigai32b9a1fd-a351-8064-9375-dc9a8f839d7a, data source collection://32b9a1fd-a351-809d-bd4d-000b0d579048README.md in ikigai workspacePROJECTS.md in ikigai workspaceCLAUDE.md (maps project names to git repo paths)initiatives/agent-team.md (PR #44) or Notion card 32c9a1fda3518134bf98c8cfa049d61cRun all sub-steps in parallel. No side effects.
Search the Control Center for all cards where Status is NOT "Done":
notion-search(query: "*", data_source_url: "collection://32b9a1fd-a351-809d-bd4d-000b0d579048", page_size: 25)
Build a dedup set: { title (lowercase) + project } for every non-Done card. Also note cards with Status = "Suggested" (stale suggestions from previous runs).
Read ~/Projects/ikigai/PROJECTS.md. For each project section, extract all unchecked - [ ] items. These are the primary candidate pool.
Read ~/Projects/ikigai/README.md. Extract:
For each project in the Project Path Registry (from CLAUDE.md), run:
cd <repo_path> && gh pr list --state open --json number,title,url,reviewDecision,isDraft 2>/dev/null
PRs needing review or with unresolved comments become task candidates.
tmux list-sessions -F "#{session_name}" 2>/dev/null | grep "^wf-"
For each wf-* session, check if it's still running or done. Done agents may have follow-up work. Failed agents may need retry.
gog cal ls --from today --to "$(date -v+7d +%Y-%m-%d)" 2>/dev/null
Flag events that create urgency (festivals, partner meetings, deadlines).
For each project repo in the Project Path Registry:
cd <repo_path> && git status --short 2>/dev/null
Repos with uncommitted changes = potential "commit and PR" candidates.
For each candidate task from Phase 1, assign a priority:
Scoring factors:
Priority mapping:
Route each task to the right agent:
| Signal | Agent | |--------|-------| | Code changes, PR reviews, bug fixes, architecture, tests, deployments | Viktor | | Blog posts, SEO audits, social media, content creation, marketing campaigns, YouTube | Luna | | Strategy docs, OKR analysis, business models, market research, pitch materials | Marco | | Contact enrichment, event follow-up, community outreach, partnership ops | Kai | | Inbox processing, calendar sync, skill maintenance, project sync | Maya (self — not dispatched) |
When in doubt, check:
For each candidate, before creating a Notion card:
wf-* tmux session is working on a related task for the same project → skip.For each surviving candidate, create a Notion card:
notion-create-pages(
parent: { data_source_id: "32b9a1fd-a351-809d-bd4d-000b0d579048" },
pages: [{
properties: {
"Name": "<task title>",
"Status": "Suggested",
"Project": "<project name from dropdown>",
"GTD Type": "Action",
"Priority": "<🔴 High | 🟡 Medium | 🟢 Low>",
"Source": "Daily Review",
"Effort": "<S | M | L>",
"Assigned Agent": "<Viktor | Luna | Marco | Kai | Maya>"
},
content: "## S3 Analysis\n\n### Tension\n<1-2 sentences: what gap exists between current state and desired state?>\n\n### Driver\n| Conditions | Effect | Relevance |\n|---|---|---|\n| <observable facts> | <consequences if unaddressed> | <connection to OKRs or mission> |\n\n### Requirement\n> <who> needs <what conditions> so that <desired outcomes>\n\n### Response Options\n- [ ] <the suggested action — what the agent will do>\n- [ ] <alternative approach>\n- [ ] Defer to next week\n\n---\n\n**Source:** <where this task came from — e.g., PROJECTS.md → sdtv → 'Fix Dropbox sync in prod'>\n**OKR Alignment:** <O1: KR3 | O2: KR1 | None>",
icon: "💡"
}]
)
Batch creation: Create up to 10 cards per run. If more candidates exist, pick the top 10 by priority and note the overflow in the summary.
Output to chat:
## Suggested Agent Tasks
| # | Task | Project | Agent | Priority | Effort | OKR |
|---|------|---------|-------|----------|--------|-----|
| 1 | ... | ... | Viktor | 🔴 High | S | O1:KR1 |
| 2 | ... | ... | Luna | 🟡 Medium | M | O2:KR1 |
Created N new "Suggested" cards in Notion Control Center.
Skipped M tasks (already tracked in Notion).
Stale suggestions (>7 days): K cards — consider archiving.
**Next:** Review in Notion → move approved cards to "To do" → run `/dispatch-approved`
tools
--- name: handoff description: Get an agent past a browser/UI wall it can't (or must not) cross on its own — a login-gated dashboard, a CAPTCHA, a 2FA prompt, an API that keeps rejecting the write, or an irreversible click that policy says a human must make. This skill is an ESCALATION LADDER, not a first move: it tells you to try the automated browser surfaces FIRST (Chrome-in-Claude, computer-use, an autonomous browser sub-agent) and only fall back to the Handoff app — a wrapper browser that h
documentation
Summarize one or more YouTube videos from their links. Use this whenever the user pastes a youtube.com or youtu.be URL (or several) and wants to know what it's about — phrasings like "summarize to telegram", "tldr these videos", "what do these say", "summary of this talk", or just dropping links with no instruction at all. Fetches each video's real transcript via yt-dlp (not the page text, which never contains the transcript), cleans the captions, and writes a per-video summary. Default delivery is Telegram; honor any other surface the user names ("to my notes", "just here in chat", "email it"). Trigger even when the user only pastes bare links — bare YouTube links almost always mean "tell me what's in these".
data-ai
Daily Digest — Chief-of-Staff role consolidates the six top-managers into one Telegram message to the Commander, instead of six. Implements the protocol from agent-proactivity.md.
development
Seed a new or empty Instagram account with a 9-post grid (3×3) so the profile looks established the moment a new visitor lands. Designed for festivals, new businesses, product launches, conferences, communities — any time an empty IG profile would hurt conversion from external traffic (QR scans, flyer drops, cross-promo). Generates assets via /image-from-gemini (per content-publishing rules — never HTML), writes captions with hashtag sets, and outputs a posting order + cadence plan. Trigger generously: phrases like '9 posts for instagram', 'fill my IG', 'starter grid', 'launch grid', 'instagram seed', '9-post grid', 'IG account not to look empty', 'first instagram posts', 'feed bootstrap', '3x3 grid', 'instagram launch content'. Even if the user mentions only one piece (just the images, just the captions, just the order), use this skill — the grid only works as an integrated bundle.