agentleFS
alabenkhlifaalabenkhlifa/dayjob/skills/weekly-update/SKILL.md

weekly-update

Skill1 starsChanged 30 days agoMITView on GitHub

Use when the user wants a weekly status update, end-of-week report, or "what did I get done this week" summary for their manager or team.

7 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 7

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run these; skip any command that fails and keep going.

    line 12
  2. Very specific:

    If today is Monday (date +%u prints 1) and the user did not say "this week", use last week instead.

    line 21
  3. Very specific:

    If that is empty, retry with --author="<name>".

    line 27
  4. Very specific:

    If <base> comes back empty, skip the not-on-base check.

    line 39
  5. Distinctive:

    If a board tool is available (Jira, Linear, ClickUp, Trello, Asana, GitHub or GitLab issues, as MCP tools): my tickets moved this week.

    line 40
  6. Distinctive:

    Risks: only real ones: failing CI, a PR stuck in review, something the user said.

    line 48
  7. Distinctive:

    Do not run tests or other project code.

    line 52
  8. Distinctive:

    Every bullet is backed by a commit, PR, ticket, or the user's own words.

    line 76
  9. Very specific:

    Never include work by other authors.

    line 77
  10. Distinctive:

    Never show tool errors in the update.

    line 79
  11. Distinctive:

    Never send anything without an explicit yes.

    line 80

“Common” is measured: kept by at least 6 in 10 of the 6 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

Weekly update

Draft a paste-ready weekly update from real activity: Shipped / In progress / Risks / Asks.

Gather context

Run these; skip any command that fails and keep going.

  1. Who am I: git config user.name and git config user.email.
  2. Window: this week, Monday 00:00 to now. Get the dates:
    mon=$(date -v-$(( $(date +%u) - 1 ))d +%F 2>/dev/null || date -d "$(( $(date +%u) - 1 )) days ago" +%F)
    prev=$(date -v-$(( $(date +%u) + 6 ))d +%F 2>/dev/null || date -d "$(( $(date +%u) + 6 )) days ago" +%F)
    • This week (default): --since="$mon 00:00"; PRs updated:>=$mon.
    • If today is Monday (date +%u prints 1) and the user did not say "this week", use last week instead.
    • Last week: --since="$prev 00:00" --until="$mon 00:00"; PRs updated:$prev..$mon.
    • A range the user gives wins ("this week", "last week", "last 7 days", dates).
  3. Audience: the manager unless the user says the team. Ask only if it changes the content.
  4. My commits on all branches: git log --all --no-merges --author="<email>" --since="<window start>" --format='%h %ad %s' --date=short If that is empty, retry with --author="<name>". If user.email and user.name are both empty, ask the user which author to use; never run git log with an empty --author.
  5. My PRs, only with a GitHub remote. The user may have several gh accounts, and the active one may not see this repo. List each account's access:
    for u in $(gh auth status 2>&1 | sed -n 's/.*account \([^ ]*\).*/\1/p'); do echo "$u $(GH_TOKEN=$(gh auth token --user "$u") gh repo view --json viewerPermission -q .viewerPermission 2>/dev/null)"; done; true
    Pick the first account with ADMIN, MAINTAIN or WRITE, else one with READ or TRIAGE. If none shows a permission, skip every gh call. Run every gh call as that account, GH_TOKEN=$(gh auth token --user <account>) gh ..., with 2>/dev/null. Never run gh auth switch. gh pr list --author @me --state all --search "updated:>=<YYYY-MM-DD>" --limit 30
  6. Work in progress: git status --short --branch and branches with my recent commits that are not on the base branch (<base>, via $base):
    base=$(git symbolic-ref --short refs/remotes/origin/HEAD 2>/dev/null | sed 's|^origin/||')
    [ -n "$base" ] || base=$(git branch --list main master | head -1 | tr -d ' *')
    If <base> comes back empty, skip the not-on-base check.
  7. If a board tool is available (Jira, Linear, ClickUp, Trello, Asana, GitHub or GitLab issues, as MCP tools): my tickets moved this week. Otherwise skip.

If steps 4, 5 and 7 find nothing for this identity, reply: "No activity found for <email> in the last 7 days." Then ask what to include. Never fill the gap with other people's work, and never name or describe their commits, even to say you left them out.

Write

  • Shipped: outcomes, not activity. "Customers can check out from the cart page", not "added checkout.js". Group related commits and PRs into one bullet.
  • In progress: each item with its next step, and an ETA only if you know one.
  • Risks: only real ones: failing CI, a PR stuck in review, something the user said. Otherwise "None".
  • Asks: decisions or help the user needs from the reader. Otherwise "None".
  • Manager audience: lead with impact, keep it short. Team audience: a bit more technical detail is fine.
  • Plain language. No commit hashes, no feat:/fix: prefixes.
  • Do not run tests or other project code.

Output template

**Weekly update — <Mon DD> to <Mon DD>**

**Shipped**
- <outcome>

**In progress**
- <item> — <next step>

**Risks**
- None

**Asks**
- None

Max 4 bullets per section. If a Slack, email or ticket tool is available, end with "Want me to send it?" and send only after an explicit yes. Otherwise end after the template.

Rules

  • Every bullet is backed by a commit, PR, ticket, or the user's own words.
  • Never include work by other authors. Always pass --author to git log.
  • Normal case: the reply is the update and the closing line. No notes about sources, skipped commits, other authors, or how you gathered data.
  • Never show tool errors in the update; leave that source out.
  • Never send anything without an explicit yes.

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.