agentleFS
CharlesLaderaCharlesLadera/claude-custom-skills/start-of-day/SKILL.md

start-of-day

Skill1 starsChanged 21 days agoMITView on GitHub

Use when starting a work session to restore context. Lists per-session journals (docs/custom-agents/memory/sessions/) so you can pick which thread to resume; supports focusing one session, including finished ones, or falling back to the legacy dated journals.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read $CFG if it exists: taskFilesDir (default docs/custom-agents/tasks), project (the folder under it), worktreeDir (default .worktrees), …

    line 31
  2. Very specific:

    If files exist: read each file's frontmatter (session, status, branch, pr, last_updated) and its latest ### 🎯 Next Session Plan bullet.

    line 46
  3. Very specific:

    If gh or pr: is unavailable, fall back to the tracker (the task file's url/id), else skip that one cleanly.

    line 61
  4. Distinctive:

    Wait for the user to choose — do not auto-open a session.

    line 77

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Start of Day Brief

Overview

Sessions are per-thread journals at docs/custom-agents/memory/sessions/<slug>.md (written by /end-of-day). With no argument this lists the active sessions and asks which to focus — it does not auto-open one. Legacy work-journal-YYYY-MM-DD.md files stay readable via the legacy argument.

Argument Reference

ArgumentBehavior
(none)List active + paused sessions; ask which to focus
<slug>Open that session's journal (full history, latest entry highlighted)
allList sessions including done ones
legacyLegacy dated-journal briefing (most recent file)
legacy <n | date | last-week>Legacy briefing with the old count / ISO-date / last-week filter

(all and legacy are reserved words; a session should not be named either.)

Steps

0. Resolve ticket-workflow conventions

Find the config the way start-ticket step 0 does — the main checkout's copy first:

HUB=$(git worktree list --porcelain | sed -n 's/^worktree //p' | head -1)   # main checkout is always first
CFG="$HUB/docs/custom-agents/ticket-workflow.config.json"
[ -f "$CFG" ] || CFG="$(git rev-parse --show-toplevel)/docs/custom-agents/ticket-workflow.config.json"

Read $CFG if it exists: taskFilesDir (default docs/custom-agents/tasks), project (the folder under it), worktreeDir (default .worktrees), contextBranch. If the file is absent, use the defaults; if $ROOT/<taskFilesDir> (ROOT below) doesn't exist either, skip step 2.5 entirely rather than searching the repo for task files.

Then the context root every step below reads, per start-ticket's Context home rule: with contextBranch set, CA="$HUB/docs/custom-agents" (and ROOT="$HUB" for repo-relative keys like taskFilesDir) — verify the hub is on contextBranch and STOP if not; in local mode (folder gitignored), the same CA/ROOT with no branch check; otherwise CA="$(pwd)/docs/custom-agents", ROOT="$(pwd)", exactly as before.

1. Parse the argument

  • none → list mode (active + paused).
  • all → list mode, including done.
  • legacy [...] → legacy daily briefing (step 5).
  • anything else → treat as a <slug>; if it matches a sessions/*.md, focus it (step 4), otherwise show the list and note no such session.

2. List session files

ls "$CA/memory/sessions/"*.md 2>/dev/null | sort
  • If none found: tell the user there are no per-session journals yet; suggest /end-of-day to create one, and mention /start-of-day legacy if old dated journals exist.
  • If files exist: read each file's frontmatter (session, status, branch, pr, last_updated) and its latest ### 🎯 Next Session Plan bullet.

2.5. Reconcile ticket delivery status

Task files carry two status axes: status: (implementation — the author's work: in progress/complete/blocked) and ticket: (delivery — the PR's journey to prod: in-review → merged → done). While you were away, PRs that were in-review may have merged. Reconcile the local ticket: field before briefing so the day starts from truth:

# task files still awaiting merge (delivery axis open) — <taskFilesDir> from step 0
grep -rl "^ticket: in-review" "$ROOT/<taskFilesDir>" 2>/dev/null

For each match, read its pr: field and check the merge state:

gh pr view <pr> --json state -q '.state'   # MERGED | OPEN | CLOSED
  • MERGED → set that task file's ticket: merged (or done if the tracker shows it shipped), bump synced_at:, and offer worktree cleanup — finish-ticket deferred it until merge, so this is the moment: git worktree remove <worktreeDir>/<slug> (offer, never auto-remove; skip if the worktree has uncommitted work).
  • OPEN → leave in-review.
  • If gh or pr: is unavailable, fall back to the tracker (the task file's url/id), else skip that one cleanly.

Surface a one-line summary of anything that flipped (e.g. "ABC-42 / PR #123 merged → task file updated, offered worktree cleanup"). A task file with no ticket: field is still in implementation (pre-PR) — ignore it here; finish-ticket is what writes ticket: + pr: when it opens the PR, so pre-PR files legitimately lack them. This reconciliation operates on the task files under $ROOT — the current checkout, or the hub when contextBranch is set or local mode is on. In local mode, commit nothing. With contextBranch, commit every file it flipped to contextBranch in one commit (docs(task): reconcile delivery status), explicit paths only, and push — per start-ticket's Context home rule.

3. Present the session list (list mode)

Exclude status: done unless all was passed. Sort by last_updated, newest first:

Active sessions:

SessionStatusLast updatedBranch / PRNext up
csv-exportactive2026-07-08feat/… · #47Merge PR #47
…

Which session do you want to focus on? (all to include finished ones · legacy for the old daily journals)

Wait for the user to choose — do not auto-open a session.

4. Focus a session

For the chosen <slug>, Read sessions/<slug>.md and present its latest dated entry as the brief (earlier entries remain in the file for scrollback):

Session <slug> — [latest date] · status [status] · branch [branch] · PR [pr]

✅ What was done: [Accomplished bullets]

🔄 Still in progress: [In Progress bullets]

🎯 Planned to start with: [Next Session Plan bullets]

📝 Notes to remember: [Notes bullets]

Skip any section the entry doesn't have — a day whose only entries came from /side-note has #### + Side session [time] blocks instead of the four ✅/🔄/🎯/📝 sections. Show those verbatim under a Side sessions heading.

Then ask: "Ready to pick up here, or has anything changed?"

5. Legacy daily briefing (legacy [...])

Fall back to the original dated-journal behavior:

ls "$CA/memory/work-journal-"*.md 2>/dev/null | sort

Apply the filter: default = most recent 1; <n> = last N by count; last-week = past 7 days; ISO date (e.g. 2026-03-30) = that date forward. Read oldest → newest and present each with the ✅ / 🔄 / 🎯 / 📝 sections (skip any missing section cleanly). If no legacy files exist, say so and point to the session list.

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.