AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Docket

skill-proptermalone-proptermaltwo-docket · by PropterMalone

Daily planner — time-aware plan coordinating the user's day and Claude's day, with calendar awareness and queue integration

No reviews yet
0 installs
32 views
0.0% view→install

Install

$ agentstack add skill-proptermalone-proptermaltwo-docket

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-proptermalone-proptermaltwo-docket)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Docket? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

> INTEGRATION STUB. This skill wires Claude Code to an external tool/account that you must supply and configure: a personal planner CLI + a queue tool + calendar access (the real setup uses bespoke docket + phyllis CLIs). It ships as a working example of the integration pattern, not a turnkey feature. See docs/integrations.md for what to install and how to plug in your own credentials. Without a personal planner CLI + a queue tool + calendar access, this skill won't run — that's expected.

Generate a daily plan that assigns tasks to calendar free slots, identifies Claude autonomous work, and signals reservations to your background-queue tool. The example setup uses two bespoke local CLIs: a planner (`, called docket in the real install) and a background-queue/scheduler (, called phyllis` in the real install). Follow these steps exactly:

1. Fetch data (parallel)

Run all of these simultaneously:

  • task-manager: Query your task manager for overdue + today's tasks plus undated p1/p2 tasks (here: Google Tasks via the gws CLI; substitute your own backend — this skill only needs a task list with dates/priorities).
  • Planner CLI: Run your planner's dry-run command to get the plan output (which reads backlog.md, slush.md, queue state, and calendar internally). In the example install this is node --import tsx ~/Projects//src/cli.ts plan --dry-run. Point this at your own planner binary/script.

2. Estimate missing durations

Check fetched task-manager tasks for missing durations. For any task without a duration set:

  1. Estimate how long the task will take based on its content, description, and project context. Use these rough buckets:
  • Quick admin (email, phone call, signup): 10-15min
  • Small fix, config change, short write-up: 30min
  • Focused task (code fix, document review, research): 60min
  • Deep work (new feature, exploration, complex debugging): 120min
  • Large project session (architecture, multi-file refactor): 180min
  1. Use update-tasks (or td task update --duration) to set the duration on each task (format: "30m", "1h", "2h", etc.)
  2. Report what you estimated: "Set durations: TaskA→30m, TaskB→2h, TaskC→15m"

Skip recurring tasks — their duration is usually stable and should be set manually.

3. Enhance with task-manager data

The CLI dry-run doesn't include task-manager data (no API token configured). Merge the task-manager tasks into the plan:

  1. Parse the CLI output to see what's already scheduled from backlog + queue
  2. Add task-manager tasks that aren't duplicated by backlog items:
  • Overdue task-manager tasks → prepend to scheduled blocks or flag prominently
  • Today's task-manager tasks → fit into free slots, using their duration estimates for block sizing
  • Recurring tasks → mention in a compact section
  1. Note any task-manager/backlog duplicates (same topic appearing in both)

4. Present the plan

Render the combined plan with these sections:

# Docket — YYYY-MM-DD

## Schedule
  8:00 AM – 10:00 AM  [Block label]
    - [TODAY] [Project] Task name
    - [TODAY] [T] task-manager task

  (gap: meeting 10:00-10:30)

  10:30 AM – 12:00 PM  [Block label]
    - [THIS WEEK] Task name

## Claude Autonomous
  [Claude] Task name (Xmin, size)

## Reservations → Queue
  HH:MM – HH:MM [intensity] reason

## Queue Status
  Window: X% | Weekly: X% | Queued: N tasks

## Unscheduled (N)
  - [TIER] [Project] Task name
  ... top 15, then count of remainder

## task-manager Recurring
  - Task name — due MM/DD (recurrence)

Rules:

  • Tag sources: [T] = task-manager, [B] = backlog, no tag = from CLI output
  • Show calendar events as gaps between blocks (the CLI already accounts for them)
  • Overdue items get prominent placement at the top of scheduled blocks
  • Keep it scannable — no full descriptions unless there's a blocker to flag
  • Empty sections: omit entirely
  • After rendering, ask: "Want me to write this plan? (writes plan JSON + queue reservations to the planner's state dir, e.g. ~/.docket/)"

5. Write plan (if approved)

Run the planner's write command (the example: node --import tsx ~/Projects//src/cli.ts plan, without --dry-run) to write:

  • ~/.docket/plan-YYYY-MM-DD.json — full plan
  • ~/.docket/reservations.json — signals to the queue tool
  • ~/.docket/cache/tasks-YYYY-MM-DD.json — task-manager snapshot

Also cache the task-manager data fetched via MCP by writing it to the cache file, so the CLI can replan later without MCP.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.