iterative-retrieval
Pattern for progressively refining context retrieval to solve the subagent context problem
Install with Tessl CLI
npx tessl i github:affaan-m/everything-claude-code --skill iterative-retrievalOverall
score
60%
Does it follow best practices?
If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:
npx tessl skill review --optimize ./path/to/skillValidation for skill structure
Discovery
7%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This description is too abstract and technical to be useful for skill selection. It describes a conceptual 'pattern' without specifying concrete actions Claude can perform, lacks natural trigger terms users would say, and provides no guidance on when to use this skill.
Suggestions
Replace abstract 'pattern' language with specific actions (e.g., 'Retrieves and refines context for subagent tasks by querying knowledge bases, filtering results, and iteratively narrowing scope').
Add a 'Use when...' clause with natural trigger terms (e.g., 'Use when working with multi-agent systems, context windows, or when subagents need targeted information retrieval').
Define the problem domain in user-friendly terms rather than technical jargon like 'subagent context problem'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses abstract language like 'pattern' and 'progressively refining context retrieval' without listing any concrete actions. It describes a concept rather than specific capabilities. | 1 / 3 |
Completeness | Missing both clear 'what' (no concrete actions listed) and 'when' (no explicit trigger guidance or 'Use when...' clause). The description only vaguely hints at a problem domain. | 1 / 3 |
Trigger Term Quality | Contains technical jargon ('subagent context problem', 'context retrieval') that users would not naturally say. No common user-facing keywords are present. | 1 / 3 |
Distinctiveness Conflict Risk | The term 'subagent context problem' is somewhat specific to a niche domain, which provides some distinctiveness, but the vague 'pattern' framing could overlap with other architectural or design pattern skills. | 2 / 3 |
Total | 5 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-crafted skill that clearly explains an iterative retrieval pattern with executable code examples and practical demonstrations. The workflow is exceptionally clear with explicit phases and stopping conditions. Main weaknesses are some verbosity in the problem statement section and the content being somewhat monolithic rather than split across files for a skill of this length (~150 lines).
Suggestions
Trim the 'Problem' section - Claude understands subagent context limitations without detailed explanation of what they don't know
Consider moving the practical examples to a separate EXAMPLES.md file and referencing it from the main skill
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is reasonably efficient but includes some verbose explanations (like the problem statement section explaining what subagents don't know) that Claude would already understand. The code examples are helpful but some surrounding text could be tightened. | 2 / 3 |
Actionability | Provides fully executable JavaScript code examples for each phase, concrete scoring criteria (0.8-1.0, 0.5-0.7, etc.), and practical examples showing the pattern in action. The integration section gives copy-paste ready prompt text. | 3 / 3 |
Workflow Clarity | Excellent workflow clarity with a clear 4-phase cycle (DISPATCH → EVALUATE → REFINE → LOOP), explicit stopping conditions ('3 high-relevance files'), and the loop includes validation logic (hasCriticalGaps check). The ASCII diagram and numbered phases make the sequence unambiguous. | 3 / 3 |
Progressive Disclosure | The content is well-structured with clear sections, but it's somewhat monolithic - the practical examples and best practices could potentially be in separate files. The 'Related' section references external resources but the main content is all inline rather than appropriately split. | 2 / 3 |
Total | 10 / 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.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- 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.