CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

schedule-ops-skill

当用户要把任务交给后台异步执行、按 cron / 周期性重复、委派给项目 Agent 长时间跑,或查看/审批/取消看板任务时触发。典型说法"每天 9 点帮我 X"、"让 coder agent 跑这个"、"有哪些待审批"。**不用于**:一次性即时答复(→直接用工具)、真实日历会议 / 约会(→calendar-ops-skill)、一次性计划审批(→`SubmitPlan`)。

68

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

73%Weight 40%Scale 1-5

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

The body is a well-structured, highly actionable operational guide for three scheduled-task tools: it covers tool selection, the critical wait-vs-reply decision with rationale and anti-patterns, two-phase approval semantics with skip criteria, scheduling types, and validation-gated workflows. Its weaknesses are token efficiency — a cron-syntax primer Claude doesn't need and a common-errors section that largely restates earlier sections — and placeholder-elided create examples that stop just short of copy-paste ready.

Suggestions

Cut the '常用 cron 示例' line — Claude already knows 5-field cron semantics; keep only the skill-specific guidance about mandatory timezone and the intervalMs floor.

Deduplicate '常见错误与防范': items like '对定时任务调 ScheduledTaskWait', '立即任务 create 后不调 ScheduledTaskWait', and 'Bash sleep + 轮询' restate the dedicated decision sections verbatim — fold them into one cross-referenced line each or remove them.

Show one complete `ScheduledTaskManage(action:'create', schedule:{...})` call block with the actual schedule object (cronExpr/intervalMs/scheduleAt + timezone) instead of 'schedule:{...}' placeholders, so create examples are copy-paste adaptable like the resolve example already is.

DimensionReasoningScore

Conciseness

Mostly efficient but with notable over-explanation and redundancy: the '常用 cron 示例' block teaches cron syntax Claude already knows, and the '常见错误与防范' section repeats guidance already covered in dedicated sections (e.g., '对定时任务调 ScheduledTaskWait' and 'Bash sleep + 轮询' restate the earlier decision rules nearly verbatim). Not 4 because the cron primer and the four-fold duplication are more than minor trimmable instances; not 2 because the core guidance (tool selection, wait-vs-reply decisions, skipPlanConfirm criteria) is skill-specific and earns its tokens.

3 / 5

Actionability

Concrete, mostly executable guidance: exact tool names and actions ('ScheduledTaskManage(action:\'create\', skipPlanConfirm:true)'), a full resolve call block, and specific parameter values (timeoutSec:60, intervalMs:1800000, cronExpr examples). Not 5 because several create examples use placeholder elisions ('...', 'schedule:{...}') rather than a complete copy-paste-ready call including the schedule object.

4 / 5

Workflow Clarity

Workflows are clearly sequenced with explicit validation and feedback loops: '先查后删' (query before delete) gates the batch-cleanup workflow, `Bash` date verification precedes once-task creation, and the timeout section provides an explicit decision tree (re-wait once, then hand off, capped at two waits) plus a correct-vs-wrong polling anti-pattern. The destructive/batch validation requirement is satisfied by the explicit check-first steps.

5 / 5

Progressive Disclosure

Good single-file structure: a tool inventory table, clearly headed sections with rule separators, and a logical progression from tool selection to decisions to workflows. Not 5 because at ~190 lines, reference material (cron primer, common-error catalog, resolve semantics) could be offloaded to a reference file, and the errors section's overlap with earlier sections signals an organization gap; not 3 because navigation is easy and no content is buried.

4 / 5

Total

16

/

20

Passed

Description

95%Weight 40%Scale 1-5

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

A strong description: it states concrete capabilities, includes quoted natural trigger phrases, explicitly answers both what and when, and actively disambiguates against adjacent skills with an exclusion clause. The only weakness is that the capability list is framed from the user's intent rather than as a comprehensive action inventory, leaving minor coverage gaps.

DimensionReasoningScore

Specificity

The description names several concrete capabilities: '把任务交给后台异步执行、按 cron / 周期性重复、委派给项目 Agent 长时间跑,或查看/审批/取消看板任务' (hand tasks to background async execution, cron/periodic repetition, delegate to project agents, view/approve/cancel tasks) — several specific actions with minor gaps in coverage (e.g., create/archive semantics are only implied). Not 5 because capability coverage is trigger-oriented rather than a comprehensive action list; not 3 because more than 1-2 concrete actions are explicitly stated.

4 / 5

Completeness

Both 'what' (background async execution, cron/periodic scheduling, agent delegation, view/approve/cancel of tasks) and 'when' (explicit trigger conditions with quoted example phrases, plus an explicit '**不用于**' exclusion clause routing one-off answers, calendar items, and plan approvals elsewhere) are clearly and explicitly answered with concrete trigger phrases.

5 / 5

Trigger Term Quality

It quotes natural user phrasings verbatim — '每天 9 点帮我 X', '让 coder agent 跑这个', '有哪些待审批' — covering scheduling, delegation, and status-query variations users would actually say, plus explicit negative triggers. Comprehensive natural-term coverage including synonyms across all trigger categories.

5 / 5

Distinctiveness Conflict Risk

Clear niche with minimal conflict risk: the '**不用于**' clause explicitly disambiguates against direct tool use, calendar-ops-skill, and SubmitPlan, and the quoted trigger phrases ('有哪些待审批', '让 coder agent 跑这个') are distinctive to task-scheduling semantics rather than overlapping with adjacent skills.

5 / 5

Total

19

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

itsablabla/openloaf-web

Path

apps/server/src/ai/builtin-skills/schedule-ops/SKILL.md

Reviewed

Table of Contents

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.