CtrlK
BlogDocsLog inGet started
Tessl Logo

iterative-retrieval

为了解决子智能体(Sub-agent)的上下文问题,逐步优化上下文获取的模式

51

Quality

56%

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

Fix and improve this skill with Tessl

tessl review fix ./docs/ja-JP/skills/iterative-retrieval/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

62%Weight 40%Scale 1-3

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

The body presents a clear, well-sequenced iterative-retrieval workflow with concrete checkpoint and feedback logic, but its code relies on undefined helpers and the material is packed into one long file without progressive file structure. It is actionable at the pattern level but not copy-paste ready.

Suggestions

Define or stub the helper functions (retrieveFiles, scoreRelevance, extractPatterns, mergeContext, hasCriticalGaps) so the examples are executable rather than pseudocode.

Move the two full case studies into a separate references/ file and link to it from the main body to reduce inline length and improve progressive disclosure.

Trim the problem section since the description already establishes the motivation, keeping only the failure modes that directly motivate the iterative approach.

DimensionReasoningScore

Conciseness

The body is mostly efficient with lean code blocks, but the re-stated problem section and two full case studies add length that could be tightened without losing clarity.

2 / 3

Actionability

Concrete structured code is provided for all four phases, but it depends on undefined helpers (retrieveFiles, scoreRelevance, extractPatterns, mergeContext, etc.), making it pseudocode rather than copy-paste executable.

2 / 3

Workflow Clarity

The four-phase DISPATCH→EVALUATE→REFINE→LOOP sequence is clearly diagrammed with an explicit EVALUATE checkpoint, a refine feedback loop, and a concrete stopping condition (relevance ≥ 0.7 and no critical gaps).

3 / 3

Progressive Disclosure

The content is well-organized into sections, but it is a monolithic ~200-line single file with no bundle files; the two case studies and extensive code could be split out, and the external pointers (URL, other skills) are not clearly signaled references.

2 / 3

Total

9

/

12

Passed

Description

50%Weight 40%Scale 1-3

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 conveys a clear domain and directional purpose but stays abstract, omitting concrete actions and any explicit when-to-use trigger. It is a competent-but-generic statement rather than a precise, trigger-rich description.

Suggestions

Add a 'Use when…' clause naming concrete triggering situations, e.g. when a sub-agent lacks the context needed to start a task.

Replace the abstract phrase with concrete capabilities, e.g. 'dispatches broad queries, evaluates file relevance, and refines retrieval criteria over up to 3 cycles'.

Include natural user-facing trigger terms (e.g. 'sub-agent context', 'iterative retrieval', 'what context does my agent need') to improve trigger-term coverage.

DimensionReasoningScore

Specificity

It names the domain (子智能体上下文问题) and a high-level action ("逐步优化上下文获取"), but the action is abstract and no concrete capabilities are listed, so it does not reach the multi-specific-action level of 3.

2 / 3

Completeness

It states what the skill does (solve sub-agent context issues via progressive retrieval), but there is no "Use when…" or equivalent explicit when-to-use guidance, which caps completeness at 2 per the rubric.

2 / 3

Trigger Term Quality

Relevant keywords like 子智能体 (sub-agent) and 上下文 (context) are present, but natural trigger phrasing and common variations a user would actually say are missing.

2 / 3

Distinctiveness Conflict Risk

The sub-agent context-retrieval niche is somewhat specific, but with no explicit triggers it could still overlap with other retrieval or context-management skills.

2 / 3

Total

8

/

12

Passed

Validation

100%

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

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
xu-xiang/everything-claude-code-zh
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.