agentleFS
kratoczkratocz/skills/skills/work-end/SKILL.md

work-end

Skill1 starsChanged 50 days agoMITView on GitHub

End-of-day summary — what got done, what carries over, what's new since /work-start. Use when the user says "/work-end", "konec dne", "shrnutí dne", "wrap up the day". Does not stop the time tracking timer — that is tracker-stop.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If missing: stop with this message — in both languages.

    line 19
  2. Very specific:

    If schema_version is not 1: warn "Snapshot je z jiné verze skillu.

    line 20
  3. Very specific:

    If age_hours > 24: warn "Snapshot je z předchozího dne (před <X> hodinami).

    line 26
  4. Very specific:

    If completed_total == 0 and carry_over == 0 and new_unhandled == 0: render "Žádná aktivita dnes.

    line 90
  5. Very specific:

    Check if the session-log plugin is installed by reading ~/.claude/plugins/session-log/config.json.

    line 94
  6. Very specific:

    Read the config to find the session log directory (session_log_dir field, or default ~/Documents/claude-sessions/).

    line 97
  7. Distinctive:

    Ask "Uložit dnešní souhrn do session logu?" and offer the options:

    line 98
  8. Distinctive:

    Check whether it exists. If it does, append — a plain file write would overwrite the day's earlier entries, so shell out:

    line 103
  9. Distinctive:

    If the file doesn't exist, create it with just the summary.

    line 107

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Work End

End-of-day retrospective: what closed, what carries over, what arrived during the day.

Steps

  1. Read snapshot:

    Read ~/.claude/plugins/work/last-briefing.json.

    • If missing: stop with this message — in both languages; the config naming one is not read until step 2: "Žádný snapshot. Bez ranního /work-start nelze udělat end-of-day souhrn." / "No snapshot. Without a morning /work-start there is no end-of-day summary to make." Return.
    • If schema_version is not 1: warn "Snapshot je z jiné verze skillu. Pokračuju best-effort." Continue.
    • Compute age:
      date -u +%s
      date -u -j -f "%Y-%m-%dT%H:%M:%SZ" "<snapshot.timestamp>" +%s  # macOS
      If age_hours > 24: warn "Snapshot je z předchozího dne (před <X> hodinami). Souhrn může být zavádějící."
  2. Read effective config — same as /work-start step 1. Stop if missing.
  3. Re-fetch all enabled sources:

    Same as /work-start step 3 (parallel fetch of all enabled and available sources), BUT also include completed items for the "completed today" computation:

    • Todoist (in addition to open tasks): call mcp_Todoist__find-completed-tasks with { "since": "<midnight today UTC>" }. Get midnight via:
      date -u -j -v0H -v0M -v0S +%Y-%m-%dT%H:%M:%SZ  # macOS, today UTC midnight
      # Linux: date -u -d 'today 00:00' +%Y-%m-%dT%H:%M:%SZ
    • GitHub: in addition to open queries, run:
      • mcp__github__search_issues with { "q": "is:closed is:issue assignee:<username> closed:>=<today UTC>" }
      • mcp__github__search_pull_requests with { "q": "is:closed author:<username> closed:>=<today UTC>" }
    • ClickUp: ClickUp clickup_filter_tasks can also filter by status — call once with status=closed, date_updated_gt=<midnight epoch ms>. (Refer to the MCP tool's schema for exact arg names.)
    • Calendar: skip (events don't have a "completed today" sense).

    Build:

    • current_open — normalized open items (same as /work-start)
    • completed_today — normalized completed items closed/completed today
  4. Compute completed / carry-over / new-unhandled:

    • completed_total = completed_today (all items that finished today, regardless of whether they were in the morning briefing)
    • completed_from_briefing = subset of completed_today whose id matches a snapshot item
    • carry_over = snapshot items whose id is in current_open (still open at end of day)
    • new_unhandled = items in current_open whose id is NOT in snapshot AND NOT in completed_today

    Sort each list by score descending (re-score carry_over and new_unhandled using current data via /work-start step 5 scoring; for completed_today use the score they had in the snapshot if available, else 0).

  5. Render summary in the configured language. Every quoted user-facing string in this skill is an example written in cs — render it in effective_config.language, keeping proper nouns, IDs, paths and durations unchanged. The step-1 message above is the exception: it carries both languages because it fires before the config is read.

    ## 📊 Souhrn dne — <today's date>
    
    Doba od /work-start: **<X>h <Y>min**
    
    ### ✅ Dokončeno: <N> celkem (<M> z ranního briefingu)
    
    1. **<title>** — <source>
    ...
    
    ### 📝 Přechází na zítra: <N>
    
    1. **<title>** — <source>, score <N>
    ...
    
    ### ⚠️ Nové během dne, neřešeno: <N>
    
    1. **<title>** — <source>, score <N>, dorazilo <H>h zpět
    ...
    
    <Warnings if any>

    If completed_total == 0 and carry_over == 0 and new_unhandled == 0: render "Žádná aktivita dnes. Vše uzavřeno před snapshotem."

  6. Optionally save to session log:

    Check if the session-log plugin is installed by reading ~/.claude/plugins/session-log/config.json. If it doesn't exist, skip this step silently.

    If it exists:

    • Read the config to find the session log directory (session_log_dir field, or default ~/Documents/claude-sessions/).
    • Ask "Uložit dnešní souhrn do session logu?" and offer the options:
      • "Ano (append k dnešnímu logu)"
      • "Ne, jen vypsat"
    • If "Ano":
      • Determine today's log file name (convention: <dir>/YYYY-MM-DD.md).
      • Check whether it exists. If it does, append — a plain file write would overwrite the day's earlier entries, so shell out:
        printf '\n\n## work-end summary\n\n%s\n' '<rendered summary above>' >> '<path>'
      • If the file doesn't exist, create it with just the summary.

    Confirm: "Uloženo do <path>."

Edge cases

  • No sources fetched successfully: render only "Žádný zdroj nedostupný. Souhrn dne nelze sestavit. Zkontroluj MCP servery." Skip lists.
  • All snapshot items still open, nothing completed today: render with completed_total = 0 section saying "(žádné položky dokončeny dnes)", and full carry_over list.
  • Snapshot from prior day: render but with warning about age. Don't try to "extend" the snapshot range — be honest that the comparison baseline is stale.

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.