miloco-create-task

创建/管理"持续运转的家庭任务"—— 由 rule（条件触发自动化）/ schedule（定时提醒）/ record（行为累积统计）/ lifecycle（长期/限时）组合装配；也是 [感知引擎] 语音指令类系统消息的统一入口。覆盖：定时/累积提醒、"X 时 Y" 自动化、行为统计、持续状态阈值等。

Quality

68%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/skills/miloco-create-task/SKILL.md

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The skill body is highly actionable with a clear multi-phase workflow and strong validation/rollback feedback loops, but it is a monolithic wall of text that keeps the entire create-path detail inline rather than progressively disclosing it. Minor repetition of the output-protocol reload banner slightly hurts token efficiency.

Suggestions

Factor the dense dimension判据 (Rule/Schedule/Record/Lifecycle rules and the装配示例) into one-level-deep reference files (e.g. references/rule-dimensions.md, references/assembly-examples.md) and keep SKILL.md as a routed overview.

De-duplicate the '[输出协议 reload]：…' banner — state the output protocol once and reference it instead of repeating the full block at each section boundary.

Consider a compact decision-tree or summary table at the top of the create path so the high-level flow is scannable before the detailed sub-dimension rules.

Dimension	Reasoning	Score
Conciseness	The body is information-dense and avoids explaining concepts Claude already knows, but it runs ~840 lines and repeats the verbatim '[输出协议 reload]：…' banner at multiple section boundaries, so it is mostly efficient but could be tightened.	2 / 3
Actionability	It provides concrete executable miloco-cli commands, exact Rule-flag and Record-content JSON mappings, and nine full copy-paste bash examples, matching 'fully executable code/commands; specific examples; copy-paste ready'.	3 / 3
Workflow Clarity	The create path is clearly sequenced (前置检查 → 第一层维度判 → 第二层子维度 → 第三层装配映射) with explicit validation checkpoints: a post-assembly 装配后自检 checklist, CLI报错后重审全命令 re-audit, and 装配失败回滚 — destructive/batch ops have feedback loops, so not capped at 2.	3 / 3
Progressive Disclosure	Only the CRUD path is split into references/crud-ops.md (one level deep, properly linked, real file); the entire create path — detailed Rule/Schedule/Record 判据 and all examples — is inlined as a ~840-line wall, matching 'content that should be separate is inline'.	2 / 3
	Total	10 / 12 Passed

Description

60%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description concretely names the skill's domain and component assembly with good action specificity, but blends natural trigger terms with internal jargon and lacks an explicit 'Use when' clause. Distinctiveness is moderate due to generic reminder-style trigger phrasing.

Suggestions

Add an explicit trigger clause (e.g. 'Use when the user asks to set up or manage recurring home tasks — timed/cumulative reminders, X-then-Y automations, behavior counting, persistent state thresholds').

Move technical component names (rule/schedule/record/lifecycle) behind plain-language phrasing or relegate them to the body so the description reads in natural user terms.

Sharpen trigger terms with the concrete phrases users actually say (e.g. '每天提醒我喝水', '客厅没人关灯', '久坐提醒') to improve both trigger quality and distinctiveness.

Dimension	Reasoning	Score
Specificity	Description lists multiple concrete actions/components — "创建/管理持续运转的家庭任务", assembly from "rule / schedule / record / lifecycle", and coverage of "定时/累积提醒、'X 时 Y' 自动化、行为统计、持续状态阈值" — matching the 'lists multiple specific concrete actions' anchor.	3 / 3
Completeness	It clearly answers 'what' (create/manage home tasks via four-component assembly) but 'when' is only implied via the 覆盖 list with no explicit 'Use when…' trigger clause, which the guidelines cap at 2.	2 / 3
Trigger Term Quality	It mixes natural terms users would say (定时提醒, 累积提醒, 行为统计) with internal technical jargon (rule/schedule/record/lifecycle 组合装配, [感知引擎] 语音指令), so coverage is relevant but missing common variations and partly jargon-laden.	2 / 3
Distinctiveness Conflict Risk	The home-task-assembly niche is fairly specific, but generic trigger terms like 定时提醒 / 累积提醒 could still overlap with general reminder skills, matching 'somewhat specific but could still overlap'.	2 / 3
	Total	9 / 12 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (850 lines); consider splitting into references/ and linking	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: XiaoMi/xiaomi-miloco
Path: plugins/skills/miloco-create-task/SKILL.md
Commit: f156d34

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.