agentsclimarketplace
Sign in

Dida task assistant

Skill jinweilong1990/dida-task-assistant-skill/dida-task-assistant

Manage Dida365 tasks while preserving a local-first record of tasks, reminders, completed work, notes, and ideas. Use when the user asks an AI agent to capture a thought, add or update a reminder or task in 滴答清单/滴答/Dida365, turn a spoken plan into action items, record completed work, add subtasks or tags, inspect task lists, or decide whether an idea should stay a local note rather than becoming a task.From its SKILL.md

Install
npx -y skills add jinweilong1990/dida-task-assistant-skill --skill dida-task-assistant

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

4.8 KB, 989 tokens by cl100k_base, as published. Nobody here has run it

Dida Task Assistant

Treat the local record as the primary capture layer and Dida365 as the task-execution connector. Do not describe this as a background two-way sync service.

Product intent and boundaries

  • Treat fragmented thoughts from Dida365, Feishu, phone notes, and chat as one natural-language capture stream. Preserve the original thought locally before deciding whether it becomes an action.
  • Turn explicit actions into tasks or reminders; keep ideas and background facts as local notes unless the user requests follow-up work.
  • Accept input from Codex, Claude Code, or another compatible local Agent Skills client. Also accept text transcribed by an upstream WeChat or voice workflow.
  • Never imply that this package includes a WeChat bot, WorkBuddy/Claw bridge, or speech recognition. Treat those as host integrations.
  • Describe additional connectors, continuous remote polling, conflict reconciliation, deterministic global merging, and background two-way synchronization as future directions, not shipped capabilities.

Resolve SKILL_DIR as the absolute directory containing this SKILL.md. Use absolute paths derived from SKILL_DIR when running bundled scripts; do not assume the current working directory is the Skill directory. The command examples use POSIX shell notation; adapt path quoting to the host shell when necessary.

First use

Before any Dida365 API operation, run python3 "$SKILL_DIR/scripts/configure.py" --show.

  • If it reports missing credentials, tell the user to create their own application at https://developer.dida365.com/, set the redirect URL to http://localhost:8080/callback, then run python3 "$SKILL_DIR/scripts/configure.py" and python3 "$SKILL_DIR/scripts/auth.py".
  • Never ask for, display, save in the repository, or reuse another person's Client Secret, access token, or refresh token.
  • Read references/dida-oauth-setup.md if the user needs configuration help.

Local-first workflow

  1. Classify the input as task, reminder, completion, idea, or note.

  2. Preserve the user's original intent and wording. Create a concise title, a short optional description, and at most two inferred tags. Preserve any explicit user tags exactly.

  3. For a likely duplicate or update, inspect the relevant Dida365 list first when authorization exists. Do not use empty task titles as matching anchors.

  4. Write a local record before changing Dida365:

    python3 "$SKILL_DIR/scripts/memory.py" capture --kind task --title "..." --raw "..." --tag "..."

    Read record_id from the JSON response.

  5. Only for an explicit actionable task or reminder, call scripts/task.py create or update. Then write the returned task_id and project_id back with scripts/memory.py link.

  6. If remote sync fails, keep the local record as pending, state that the thought was retained locally, and give the user the next action. Do not silently discard or retry a destructive action.

Routing rules

  • A clear future action becomes a task or reminder.
  • A completed action becomes a local completion record. Create and immediately complete a Dida365 task only when the user asks for that external log.
  • An idea, preference, background fact, or reflection remains local unless the user asks for follow-up work. Ambiguous ideas may be labelled 灵感 or 待验证; do not force them into urgent tasks.
  • For multiple explicit steps, use --items-json to create a checklist or update --add-subtask to extend the matching task.
  • Do not complete, uncomplete, delete, or move a remote task from an ambiguous reference. Ask for the target task when needed.

Commands

Run bundled scripts through the resolved absolute Skill path:

python3 "$SKILL_DIR/scripts/task.py" project-list
python3 "$SKILL_DIR/scripts/task.py" list --project-id <id>
python3 "$SKILL_DIR/scripts/task.py" create --title "给供应商回电话" --due-date 明天 --priority 5 --tag 工作
python3 "$SKILL_DIR/scripts/task.py" update --task-id <id> --project-id <id> --add-subtask "确认报价"
python3 "$SKILL_DIR/scripts/task.py" complete --task-id <id> --project-id <id>

Use scripts/memory.py list to inspect local records without accessing Dida365. Scripts return JSON; use their IDs rather than guessing.

Project identity

This public Skill is created and maintained by Senmu. Preserve the LICENSE file and its Required Notice when redistributing permitted copies or modifications.

What ships with it: 7 files

35.2 KB alongside SKILL.md, 6 of them executable

references/

scripts/

Keep looking

Skills are one crate of 326,144. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.