unified-taskflow

【强制】复杂任务管理系统。 AI 必须在遇到以下情况时**先读取此 SKILL.md**，再执行任何操作： - 多文件变更、规划设计、新功能开发 - 用户说"帮我想想/重构/升级/实现" 禁止使用内置 agentic_mode 的 artifact 目录处理复杂任务，必须使用项目根目录下的 .taskflow/ 目录。

Quality

70%

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 ./.agent/skills/unified-taskflow/SKILL.md

Quality

Content

75%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-organized, mostly lean multi-phase workflow with concrete rules, explicit validation checkpoints, and good progressive disclosure to verified reference files. Main weaknesses are minor: a duplicated reference listing, orphaned script files not referenced anywhere, and slightly loose elastic-depth criteria.

Suggestions

Reference the two scripts (session-catchup.py, task-lifecycle.py) in the body — e.g., add a row to the 引用文件 table — so all bundle files are discoverable.

Collapse the 按需加载 list and 引用文件 table into a single reference section to remove the duplicate listing and save tokens.

Tighten '弹性深度' by giving a concrete discriminator (e.g., file-count or step-count threshold) for choosing simple vs. medium vs. complex execution.

Dimension	Reasoning	Score
Conciseness	The body is lean with tight bullet sections and no padding of concepts Claude already knows, but the four references are listed twice (按需加载 and 引用文件 table) and the RIPER-Core section adds tokens that could be trimmed.	4 / 5
Actionability	Gives concrete procedural guidance — specific file names (anchor.md, checkpoint.md, design.md), the .taskflow/ layout, and crisp rules (2-Action, 3-Strike) — but the schema of anchor/checkpoint files is delegated to templates and '弹性深度' leaves depth criteria somewhat loose.	4 / 5
Workflow Clarity	Clear Phase 0 → Execution → Completion sequence with explicit checkpoints (user confirmation, 2-Action updates, final Re-grounding) and feedback loops (3-Strike escalation, drift handling), though the elastic-depth tiering is loose enough to keep it from a clean 5.	4 / 5
Progressive Disclosure	Well-structured overview pointing one level deep to 4 verified reference files and 3 template assets with a navigation table, but two bundle scripts (session-catchup.py, task-lifecycle.py) are never referenced in the body and the reference list is duplicated.	4 / 5
	Total	16 / 20 Passed

Description

66%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.

The description gives clear, natural trigger phrases and explicit 'when' guidance, but states the capability abstractly and uses broad triggers that risk overlapping with general development tasks. Strengthening the 'what' with concrete actions and narrowing the triggers would improve distinctiveness.

Suggestions

Replace the generic '复杂任务管理系统' with concrete actions, e.g. '在 .taskflow/ 目录创建 anchor/checkpoint/design 文件，分阶段管理复杂任务的生命周期'.

Narrow triggers to reduce conflict with general coding — distinguish from routine refactors by emphasizing multi-phase or multi-file coordination scope.

Keep the existing natural trigger phrases but add a brief 'what' clause so Claude can distinguish this skill from baseline behavior.

Dimension	Reasoning	Score
Specificity	Names the domain ('复杂任务管理系统') and concrete trigger scenarios (multi-file changes, planning/design, refactor/upgrade/implement), but the core capability is stated abstractly rather than as concrete actions like 'creates anchor files and tracks checkpoints'.	3 / 5
Completeness	Provides an explicit 'when' via the trigger list ('遇到以下情况时先读取此 SKILL.md') and a 'what' ('复杂任务管理系统'), but the 'what' is generic and lacks concrete capability detail, keeping it below a 5.	4 / 5
Trigger Term Quality	Includes natural user phrases ('帮我想想/重构/升级/实现', '多文件变更', '规划设计') with synonyms, though a few common variations like '加功能/改需求' are missing.	4 / 5
Distinctiveness Conflict Risk	The triggers (multi-file changes, refactor/upgrade/implement, '帮我想想') are common enough to overlap heavily with general coding assistance and baseline behavior, creating real conflict risk.	3 / 5
	Total	14 / 20 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
relative_links	Relative link issues: 3 deeper-than-1-level	Warning
referenced_paths_exist	Referenced path issues: 3 deeper-than-1-level	Warning

	Total	14 / 16 Passed

Repository: ooooooooooooooooooop/ai_life_os
Path: .agent/skills/unified-taskflow/SKILL.md
Commit: 61ead80

Reviewed: 1 day 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.