agentleFS
luanmorenommacielluanmorenommaciel/agentspec/.claude/skills/standup-report/SKILL.md

standup-report

Skill256 starsChanged 2 months agoMITView on GitHub

Generates the user's daily standup message in the Done / Will do / Blockers format from git history, PRs/issues, and the user's own notes, ready to paste into the team chat, with an optional short follow-up thread for extra detail. Use when the user asks for "standup", "my check-in", "daily update", or to put together the day's status message for the team. Do not use for meeting summaries — use `meeting-analysis` — or for status reports longer than a standup; write those as documents.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 10 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    One line per bullet: Never hard-wrap inside a bullet — chat tools soft-wrap, and a hard-wrapped bullet renders as broken fragments.

    line 59
  2. Distinctive:

    Team @-mention is optional and never assumed: If the user wants one, ask for the exact group handle — do not invent or guess it.

    line 63
  3. Very specific:

    Git history: Run git log --oneline --since="yesterday" --author="<name>" in the current repository, plus any other repositories the user …

    line 73
  4. Distinctive:

    User-pointed sources: Read whatever task list, tracker export, or planning file the user points at.

    line 75
  5. Distinctive:

    The user's own words: Ask what the artifacts do not show: reviews given, meetings held, decisions made, teammates unblocked, direction …

    line 76
  6. Distinctive:

    Draft, then confirm: Always show the assembled draft and confirm it with the user before final delivery — artifacts under-report the day, …

    line 77
  7. Distinctive:

    When one item genuinely needs context — links, error detail, design rationale, material to attach — offer a short follow-up thread message …

    line 101
  8. Distinctive:

    Put the final message in a fenced code block, ready to select and paste into the chat.

    line 117
  9. Distinctive:

    If a thread was drafted, deliver it in its own separate fenced block after the main message.

    line 118
  10. Distinctive:

    Default is message-only: deliver the text and stop.

    line 122
  11. Distinctive:

    If the user keeps a standup history directory, offer once to save a copy there — ask for the location the first time, then reuse the answer …

    line 123
  12. Very specific:

    When saving: one file per day, named by date (e.g.

    line 124

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “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

Standup Report

Generate the user's daily standup message: what moved forward, what comes next, and what is blocked. Assemble evidence from the repository and the user's own account, draft the three-section message, confirm it with the user, and deliver it ready to paste into the team's chat tool (Slack, Teams, or whatever the team uses).

A standup is a pulse, not a report. The format is lightweight — a greeting and emoji markers are fine — but the substance stays concrete, verifiable, and honest.

When to use

  • The user asks for a standup, check-in, daily update, or the day's status message.
  • The user wants yesterday's work and today's plan summarized for the team chat.

Skip if

  • The input is a meeting transcript or the user wants a meeting summarized — use meeting-analysis instead.
  • The user needs a status update longer than a standup (weekly summary, retrospective, project report) — write those as documents, not chat messages.

Message template

<optional greeting line — e.g. "Good morning, team! Daily check-in:" — with the team @-mention if the user wants one>

🟢 **Done**
- <concrete verb + artifact, one line — e.g. "Merged PR #118 — retry handling in the sync client">
- <2–4 bullets per section>

🟡 **Will do**
- <next meaningful deliverable, one line — e.g. "Open the follow-up PR wiring retries into the batch endpoint">

🔴 **Blockers**
- <what is blocked + what is needed + from whom (by role), one line — or exactly "No blockers today.">

Example, filled:

Good morning, team! Daily check-in:

🟢 **Done**
- Merged PR #118 — retry handling in the sync client
- Opened PR #123 — pagination for the export endpoint
- Published the draft rollout plan for team review

🟡 **Will do**
- Start the auth-module migration once PR #123 lands
- Turn rollout-plan feedback into the final checklist

🔴 **Blockers**
- PR #123 (pagination for the export endpoint) awaits reviewer feedback — blocks the migration start

Formatting rules

  • One line per bullet. Never hard-wrap inside a bullet — chat tools soft-wrap, and a hard-wrapped bullet renders as broken fragments.
  • 2–4 bullets per section. More than that is a report, not a pulse — move the overflow to the thread or cut it.
  • Blank line between sections — the message breathes better.
  • The 🟢/🟡/🔴 markers and bold section headers are the format's identity — keep them. If the team's chat tool renders emoji shortcodes more reliably than literal emoji, swap them at delivery time.
  • Team @-mention is optional and never assumed. If the user wants one, ask for the exact group handle — do not invent or guess it.
  • Greeting matches the time of day, and is optional.
  • Inline links where the chat renders markdown: a labeled link to a PR or issue beats a raw URL. In Blockers, make the link label descriptive enough that the reader grasps what is blocked without clicking.
  • Joint work reads in the plural, solo work in the singular. "We aligned on the rollout plan" credits the team; "I opened the PR" owns the work. Do not claim shared work as solo, and do not dilute solo work into "we".
  • No blockers? Say so explicitly ("No blockers today.") — never invent one to fill the section.

Assemble the content

  1. Identity. Derive the author from git config user.name (and user.email for log filtering). If unset or ambiguous, ask the user. Never hardcode a person.
  2. Window. Done covers the span since the last standup — default: yesterday plus today so far. Will do covers the next working day.
  3. Git history. Run git log --oneline --since="yesterday" --author="<name>" in the current repository, plus any other repositories the user names.
  4. PRs and issues. When gh is available and the repository is on GitHub, pull the user's recently merged and open PRs (gh pr list --author "@me" --state merged / --state open) and recently closed or assigned issues (gh issue list --assignee "@me"). If gh is missing or errors, skip this source silently — it is an enrichment, not a requirement.
  5. User-pointed sources. Read whatever task list, tracker export, or planning file the user points at. Do not guess paths and do not crawl the workspace looking for one.
  6. The user's own words. Ask what the artifacts do not show: reviews given, meetings held, decisions made, teammates unblocked, direction changes.
  7. Draft, then confirm. Always show the assembled draft and confirm it with the user before final delivery — artifacts under-report the day, and the user knows what actually happened.

Quality rules

#RuleWhy
1Done bullets use concrete, verifiable verbs anchored to artifacts — "opened PR", "merged", "published", "fixed" — never "worked on" or "looked into".Vague verbs hide state.
2Verb proportional to the work: "refined" ≠ "reworked" ≠ "rebuilt" — pick the smallest verb that is still true.Inflated verbs erode trust.
3Will-do items sit at deliverable altitude — the next meaningful outcome, not micro-actions like "reply to a review comment".Teammates read outcomes, not todo lists.
4Report state at posting time: if something changed since yesterday's plan, write the current truth, not the stale intention.Stale intentions misdirect the team.
5Never cite CI or test status as an accomplishment.Green tests are table stakes, not progress.
6"Decided" / "closed" only for decisions actually made; "discussed" / "aligned" for conversations that set direction.Conflating them creates phantom commitments.
7Waiting on review or feedback IS a blocker — name it honestly, with what is needed and from whom (a role, not a personal callout).Unnamed waits stall work silently.
8An honest thin day beats a padded one.Padding is visible and costs credibility.

Two further checks before showing the draft:

  • Will do lists only what the user owns. A teammate's deliverable or a team-wide goal does not belong there; the user's own coordination step toward it ("raise the cutover date with the platform team") does.
  • Cut execution parentheticals. A detail like "(rebase expected, no conflicts)" belongs in the PR, not the pulse — if it matters, move it to the thread.

Apply the rules while drafting, not as an afterthought — most rewrites come from inflated verbs (rule 2) and missing review-wait blockers (rule 7).

Optional detail thread

  • When one item genuinely needs context — links, error detail, design rationale, material to attach — offer a short follow-up thread message instead of bloating the main message.
  • Threshold: one or two extra details belong in the bullet itself; a thread is only for genuinely dense material.
  • Thread format: free-form and short. It does not repeat the Done / Will do / Blockers structure. Title it with the date and topic.

Example thread message:

Thread — <date>: export-endpoint pagination

Context for PR #123: went with cursor-based pagination over offset — offset scans degraded badly past large row counts.
Benchmark numbers and the query-plan comparison: <link>
Client impact: none — the old offset params keep working until the next major version.

Delivery

  • Put the final message in a fenced code block, ready to select and paste into the chat. Never deliver it as a blockquote or decorated markdown — terminal border characters contaminate the copy-paste.
  • If a thread was drafted, deliver it in its own separate fenced block after the main message.

Persistence

  • Default is message-only: deliver the text and stop.
  • If the user keeps a standup history directory, offer once to save a copy there — ask for the location the first time, then reuse the answer without re-asking.
  • When saving: one file per day, named by date (e.g. YYYY-MM-DD.md), containing the message and the thread if there is one. If the day's file already exists, update it rather than duplicating.

Saved-file layout:

# Standup — YYYY-MM-DD

## Message
(the message exactly as delivered)

## Thread (optional) — "topic"
(the thread text)

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.