agentleFS
Yesterday-AIYesterday-AI/ytstack/skills/quick-task/SKILL.md

quick-task

Skill2 starsChanged 29 days agoMITView on GitHub

Quick lane for a change the user can describe in one sentence (fix, tweak, rename, bump, small refactor) in a project with .ytstack/ and PREFERENCES quick_lane on: Q###-PLAN and Q###-SUMMARY under .ytstack/quick/, a verification command, no pitch, no review, no milestone bookkeeping. Not for new features, migrations, unknown root causes or more than about three files (plan-task, systematic-debugging).

2 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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    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. Very specific:

    Use Write. Path: {YT_DIR}/quick/{NEXT_Q}-PLAN.md (create quick/ on first use).

    line 109
  2. Very specific:

    Use Write. Path: {YT_DIR}/quick/{NEXT_Q}-SUMMARY.md.

    line 139

≈ “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

quick-task

The quick lane. One sentence in, one verified change out, one Q### artifact pair as the record. No pitch, no CEO/eng review, no milestone/slice/task bookkeeping.

Anti-Pattern: "It's small, I'll just do it without any artifact"

Small changes are where "what did we change and why" evaporates first. The quick lane costs two short files and a verification command; it buys a greppable record (Q###: commits, .ytstack/quick/) and the same evidence rule as every other ytstack change. If writing the one-sentence plan feels hard, the change is not small -- route to plan-task.

Checklist

You MUST create a TodoWrite task for each of these items and complete them in order:

  1. Run preamble -- resolver, ceremony dials, next free Q###
  2. HARD-GATE -- .ytstack present, quick_lane on
  3. Fit check -- one sentence, known cause, ~3 files max; otherwise route to plan-task / systematic-debugging and stop
  4. Write Q###-PLAN.md -- sentence, files, verification command
  5. Execute -- ytstack:test-driven-development, then ytstack:verification-before-completion; commits prefixed Q###:
  6. Write Q###-SUMMARY.md -- outcome, verification output, commits
  7. Report + return -- path + one line; no auto-chaining

Process Flow

digraph quick_task {
    "Run preamble" [shape=box];
    "HAS_YTSTACK + QUICK_LANE on?" [shape=diamond];
    "Abort: init-project / plan-task" [shape=doublecircle];
    "Fits one sentence, <= ~3 files, known cause?" [shape=diamond];
    "Route to plan-task / systematic-debugging" [shape=doublecircle];
    "Write Q###-PLAN.md" [shape=box];
    "TDD + verification" [shape=box];
    "Write Q###-SUMMARY.md" [shape=box];
    "Report + return" [shape=doublecircle];

    "Run preamble" -> "HAS_YTSTACK + QUICK_LANE on?";
    "HAS_YTSTACK + QUICK_LANE on?" -> "Abort: init-project / plan-task" [label="no"];
    "HAS_YTSTACK + QUICK_LANE on?" -> "Fits one sentence, <= ~3 files, known cause?" [label="yes"];
    "Fits one sentence, <= ~3 files, known cause?" -> "Route to plan-task / systematic-debugging" [label="no"];
    "Fits one sentence, <= ~3 files, known cause?" -> "Write Q###-PLAN.md" [label="yes"];
    "Write Q###-PLAN.md" -> "TDD + verification" -> "Write Q###-SUMMARY.md" -> "Report + return";
}

Preamble

# ytstack resolver (M016): checkout = where this shell runs, via the shared lib.
_LIB="${CLAUDE_SKILL_DIR}/../../hooks/lib/git-ctx.sh"
if [ -f "$_LIB" ]; then
  . "$_LIB"; yt_git_ctx "$PWD"; yt_resolve_yt_dir >/dev/null
  _PROJECT_DIR="$YT_TOPLEVEL"; _YT_DIR="$YT_DIR"; _YT_SCOPE="$YT_SCOPE"; _IS_GIT="$YT_IS_GIT"; _PROJECT_KEY="$YT_REPO_KEY"
else
  # Host did not substitute CLAUDE_SKILL_DIR (e.g. Codex adapter): project scope only.
  _IS_GIT=$([ "$(git rev-parse --is-inside-work-tree 2>/dev/null)" = true ] && echo yes || echo no)
  _PROJECT_DIR=$(git rev-parse --show-toplevel 2>/dev/null || pwd -P)
  _YT_DIR=""; _YT_SCOPE=none; _PROJECT_KEY=""
  [ -d "$_PROJECT_DIR/.ytstack" ] && { _YT_DIR="$_PROJECT_DIR/.ytstack"; _YT_SCOPE=project; }
fi
cd "$_PROJECT_DIR" 2>/dev/null || true
_PROJECT_SLUG=$(basename "$_PROJECT_DIR")
_HAS_YTSTACK=$([ -n "$_YT_DIR" ] && echo yes || echo no)

# Ceremony dials (M018)
if command -v yt_prefs >/dev/null 2>&1; then yt_prefs "$_YT_DIR" >/dev/null; else YT_PREF_QUICK_LANE=on; YT_PREF_WARNINGS=""; fi

# Next free Q### under .ytstack/quick/
_NEXT_Q="Q001"
if [ -n "$_YT_DIR" ] && [ -d "$_YT_DIR/quick" ]; then
  _MAXQ=$(ls "$_YT_DIR"/quick/Q[0-9][0-9][0-9]-*.md 2>/dev/null | sed -E 's/.*Q([0-9]{3})-.*/\1/' | sort -u | tail -1)
  [ -n "$_MAXQ" ] && _NEXT_Q=$(printf "Q%03d" $((10#$_MAXQ + 1)))
fi
_BRANCH=$(git branch --show-current 2>/dev/null || echo unknown)

echo "HAS_YTSTACK: $_HAS_YTSTACK"
echo "QUICK_LANE: $YT_PREF_QUICK_LANE"
echo "PREF_WARNINGS: ${YT_PREF_WARNINGS:-none}"
echo "NEXT_Q: $_NEXT_Q"
echo "BRANCH: $_BRANCH"
echo "IS_GIT: $_IS_GIT"

Procedure

Step 2: HARD-GATE

  • HAS_YTSTACK=no: "quick-task needs .ytstack/. Run /ytstack:init-project (greenfield) or /ytstack:adopt-brownfield (existing codebase) first." STOP.
  • QUICK_LANE=off: "The quick lane is off in PREFERENCES.md (quick_lane: off). Use /ytstack:plan-task inside the current slice." STOP.

Step 3: Fit check (no question)

Take the user's sentence as the plan. It fits when it names one change with a known cause and touches about three files or fewer. It does not fit when it introduces a new feature or user-facing capability, changes a schema or data migration, needs investigation of an unknown root cause, or spans more files. If it does not fit, say which criterion failed and route: feature/scope -> ytstack:plan-task (or plan-milestone when no milestone is active); unknown cause -> ytstack:systematic-debugging. STOP.

Step 4: Write Q###-PLAN.md

Use Write. Path: {YT_DIR}/quick/{NEXT_Q}-PLAN.md (create quick/ on first use).

---
kind: quick
id: {NEXT_Q}
created: {ISO_TIMESTAMP}
status: in-progress
branch: {BRANCH}
---

# {NEXT_Q} -- {the user's sentence}

## Files

- {path} -- {what changes}

## Verification

```bash
{one or two real commands}

### Step 5: Execute

Invoke `ytstack:test-driven-development` with this plan as the task (RED on the verification command or a focused test, GREEN with the minimal change), then `ytstack:verification-before-completion`. Commit subjects start with `{NEXT_Q}: `.

### Step 6: Write `Q###-SUMMARY.md`

Use Write. Path: `{YT_DIR}/quick/{NEXT_Q}-SUMMARY.md`. Commits via the lib: `yt_task_commits "$_PROJECT_DIR" "^{NEXT_Q}:"` -> `YT_COMMITS`.

```markdown
---
kind: quick
id: {NEXT_Q}
closed: {ISO_TIMESTAMP}
verification: passed
---

# {NEXT_Q} -- Summary

## Commits

{YT_COMMITS}

## Outcome

{two sentences, past tense}

## Verification

Command: `{command}` -- passed. {one line of real output}

Then flip the PLAN's status: in-progress to status: done (Edit) and bump last_updated in STATE.md (no pointer change: the quick lane does not touch current_milestone / active_slice / active_task).

Step 7: Report + return

Quick task {NEXT_Q} done: {sentence}. Plan + summary under {YT_DIR}/quick/, commits {NEXT_Q}: on {BRANCH}.

Terminal State

Return control after the report. Do NOT invoke plan-task, summarize-task or reassess-roadmap: the quick lane is complete in itself and leaves the milestone pointer untouched. Aborts (HARD-GATE, fit check) return with the routing message only.

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.