agentleFS
Nicolas-RigaudyNicolas-Rigaudy/cr-0w/.claude/skills/brief/SKILL.md

brief

Skill1 starsChanged 7 days agoMITView on GitHub

Morning briefing. One screen covering reviews waiting, sprint, open MRs/PRs, open threads, vault tasks, background runs, plus mail/calendar when connected. Ends with a suggested top 3. Use when the user says "brief", "briefing", "what's on today", or starts the day.

78 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use capture, strategy and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • toneron2Skill

    capture

    Capture thoughts to inbox with automatic classification

    Also in 6 projects that use this

  • EpistatesSkill

    strategy

    Plan weekly or monthly X content strategy. Use when the user wants to plan their content calendar, set growth goals, build a content pipeline, or think strategically about their X presence.

    Also in 4 projects that use this

  • aleksander-dytkoSkill

    decision

    Make a product decision with full context from the vault and available sources

    Also in 3 projects that use this

  • dirienSkill

    decide

    Three-agent adversarial debate protocol for strategic decisions. Two advocates steelman opposing positions while a moderator identifies risks and synthesizes a binding consensus. Use when choosing between alternatives, evaluating tradeoffs, or making high-stakes decisions.

    Also in 5 projects that use this

  • JaimeGBASkill

    ingest

    Digest an external source (article, paper, transcript, URL, book chapter) and integrate its content into the knowledge base

    Also in 7 projects that use this

  • howellsSkill

    reconcile

    Audit a project against current plugin conventions and offer to update structure.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 71 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read-only except for one activity-log line.

    line 21
  2. Distinctive:

    Use the actions only as context ("reviewed !210"), and only when it changes the day (e.g.

    line 30
  3. Distinctive:

    Ask only for key, summary, status and priority fields.

    line 33
  4. Distinctive:

    Mail + calendar: only if Microsoft 365 tools are available in this session.

    line 41

“Common” is measured: kept by at least 6 in 10 of the 71 edited versions of this file we could read. “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

Brief

Runs in its own context (no conversation history): everything needed is gathered below. Return only the brief.

Pre-gathered

Everything the brief needs from files and scripts, as one JSON (bin/crow-gather brief; keys below refer to it). Don't read state files or rerun these scripts. !"${CLAUDE_PROJECT_DIR}/bin/crow-gather" brief --json 2>&1 || true

Keys: today_priorities (null unless dated today), afk (summary of a run since yesterday, log tail, report path), open_threads (Me, Waiting as "<age>d [tag] text", STALE past 14 days; other sections only counted), status, activity (crow-activity since yesterday), notes (vault notes, 3 days), releases, tidy (safe cleanups already applied, runbooks/repo-tidy.md), fleet (Logics next moves, reviews left out), long_sessions (past 120k, runbooks/long-sessions.md), accounts (headroom), usage_split (token cost split, 14 days; baseline in local/state/research/token-baseline-2026-11.md), runs (today's runs that did not succeed), vault_tasks (due on or before today, inbox count), checks (crow-checks rows: sprint rituals, 1:1s, old mail, rituals, overdue notes), retros (reports since the last brief, pending proposals, worker status), targets.activity_log. A source that failed is {"error": ...}; _cut_to_fit lists fields shortened to stay in budget.

Read-only except for one activity-log line. Target: one screen, ≤ 40 lines. Gather cheaply, in this order.

Gather

-1. Refresh before stating state: run git fetch -q in the hub and every repo you cite, and re-read MR state live (bin/crow-mrs, glab with the real HOME) before saying a branch is behind, an MR is open, or a choice is still pending. State files and the activity log lag the same morning; the owner has had to say "check again, some of it is outdated". 0. Last AFK run first: if afk.summary is set, start with how the run ended (its notes and project lines). Never state how the night ended without it; if afk.log_tail shows jobs since yesterday evening and afk.summary is null, say which jobs never finished and offer to compile the report (see runbooks/afk-runs.md, Morning).

  1. Open threads + status: open_threads and status (format: runbooks/open-threads.md). Show Me first, then Waiting with each line's age. List STALE lines and ask keep, re-date or drop (drop removes the line from the state dir's open-threads.md). ## Crow is left out on purpose. Never write a status note into the file.
  2. Forge activity: activity. Don't explore repos further. "My open MRs" lists every open MR with its unresolved thread count.
    • "My GitLab/GitHub actions" are the owner's OWN outgoing actions (comments he left on others' MRs, MRs he accepted). Never put them in "Needs you" and never write "comments from you" as a to-do: a comment he left on a colleague's MR is waiting on the colleague. Only "Awaiting my review", unresolved threads on "My open MRs", approvals pending and failed pipelines on his MRs belong in "Needs you". If all three are "none", write "Nothing waiting on you (no review, no open MR)".
    • Comments on MRs may be SonarCloud's quality-gate note posted through the owner's token (crow-activity filters them). Never read an action as a review he gave unless it is clearly his.
    • Use the actions only as context ("reviewed !210"), and only when it changes the day (e.g. a review he gave that is still unanswered after 2+ days).
  3. Tracker (optional, e.g. Jira through the Atlassian MCP; skip if none is connected): searchJiraIssuesUsingJql with assignee = currentUser() AND sprint in openSprints() AND statusCategory != Done ORDER BY priority DESC Ask only for key, summary, status and priority fields.
  4. Vault tasks: vault_tasks.due (already filtered to due on or before today) and vault_tasks.inbox_files. Don't read notes further.
  5. Release tracks: releases.
  6. Long sessions: each line of long_sessions goes under "Needs you" with the proposal: /wrap, then split the rest into crow handoff runs with one narrow end state each. Nothing listed, nothing printed. Background runs: runs.not_succeeded (running or failed). Accounts: from accounts, one line only when the recommended account is under 30% or an account the owner uses is blocked (with its reset). Checks: each row of checks.rows goes under "Needs you" as "what (why): action", the same rows as the deck's You decide. checks.skipped lists sources it could not read; say nothing about them. Logics next moves: from fleet, only moves that unblock work (blocked docs, a task at 100% in a repo in delivery); skip pure closeout chores unless nothing else is open.
  7. Retros: retros.reports (one line each: project, what was applied; open a report only when its name says nothing), retros.pending_proposals, and failures from retros.status.
  8. Mail + calendar: only if Microsoft 365 tools are available in this session. Today's meetings, and unread mail that needs a reply. If unavailable, print one line saying so and move on.
  9. Owner notes: from notes, mention anything that changes the day (a #blocage, a #decision, a note for tomorrow) in one line each. Then pick at most three live topics (open threads, sprint tickets, release tracks) with no trace in the notes or activity for 3 days and ask them in one AskUserQuestion call, only in the main session; file answers with bin/crow-note ... --section capture. Nothing stale, nothing asked.

Skip any source that errors and note it in one line at the bottom.

Output (exactly this shape, skip empty sections)

# Brief: <weekday DD/MM>

## 🎯 Your priorities today
- each line of today.md with its state (done / open), then what blocks it; skip if the file is not dated today

## 🔴 Needs you
- review feedback on my MRs (unresolved threads, always first: fixing it gets them merged), reviews waiting on me (they unblock devs), approvals pending, failed runs, overdue tasks

## 📅 Today
- meetings and recurring rituals (listed in `local/context.md`), deadlines

## 🏃 Sprint
- KEY status: summary   (mine, in priority order)

## 🧵 Open threads
- stale (>3 days) or due ones only

## 🚀 Release tracks
- one line per track: name → target, next milestone (or "no roadmap yet"); then one suggested next step for the most recently active personal project

## 🧹 Repos
- what crow-tidy cleaned (one line), then per project only: uncommitted work, unpushed main, branches older than 3 days (offer to merge/delete), worktrees in /tmp

## 🔁 Learned
- one line per retro since the last brief (what changed); pending proposals to approve; failed retros

## ✅ Suggested top 3
1. …  (follow the priority order in CLAUDE.md)

After

Append - HH:MM [hub] brief to targets.activity_log. Nothing else.

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.