CtrlK
BlogDocsLog inGet started
Tessl Logo

improving-prompts

Use when optimizing CLAUDE.md, AGENTS.md, custom commands, or skill files — diagnose the concrete failure first, then apply current documented Anthropic best practices (explicit instructions, context/motivation, examples, output and verbosity control, thinking/effort, CLAUDE.md size and skill-description rules) instead of inventing improvements. Triggers when a prompt isn't followed, a skill won't activate, CLAUDE.md is too long or ignored, or migrating prompts to current Claude models.

72

Quality

88%

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

Quality

Content

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-structured meta-skill for prompt improvement that excels in actionability and workflow clarity. Its strongest feature is the disciplined process requiring concrete failure identification before making changes, with clear go/no-go criteria. The main weaknesses are moderate verbosity (the anti-pattern table and inlined best practices could be trimmed) and the inlining of reference content that the skill itself points to an external file for.

Suggestions

Trim the 'Stop If You Catch Yourself' table to 4-5 most critical entries and move the rest to the referenced best-practices file

Reduce the inlined best practices in Step 2 to brief one-liners with a stronger pointer to `references/anthropic-best-practices.md` for details, since the skill already references that file

DimensionReasoningScore

Conciseness

The skill is mostly efficient but includes some unnecessary content. The 'Core Problem' section explains failure modes Claude could infer, and the 'Stop If You Catch Yourself' table, while useful, is somewhat verbose. The anti-pattern table has 8 entries where 4-5 would suffice. However, the quick reference table and process steps are well-condensed.

2 / 3

Actionability

The skill provides highly concrete, actionable guidance: a clear 4-step process, specific before/after examples of what counts as a concrete issue, a quick-reference issue→fix table with exact phrasing changes (e.g., 'Can you…' → imperative 'Do X'), and explicit criteria for when to proceed vs. stop. The guidance is specific enough to be directly applied.

3 / 3

Workflow Clarity

The 4-step process is clearly sequenced with explicit validation checkpoints: Step 1 requires understanding before changing (with go/no-go criteria), Step 2 references documented practices, Step 3 preserves what works, and Step 4 requires rationale for each change. The 'Stop If You Catch Yourself' table serves as an effective feedback loop for error recovery. The process has clear gates preventing premature action.

3 / 3

Progressive Disclosure

The skill references `references/anthropic-best-practices.md` for the full reference, which is good progressive disclosure. However, no bundle files are provided, so this reference is unverifiable. Additionally, the Step 2 section inlines a substantial amount of the best practices content that could arguably live entirely in the referenced file, making the main skill longer than necessary.

2 / 3

Total

10

/

12

Passed

Description

100%

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 is an excellent skill description that clearly defines its scope (optimizing Claude configuration artifacts), lists concrete actions and best-practice areas, and provides explicit trigger conditions for both proactive use and reactive troubleshooting. The dual 'Use when' and 'Triggers when' clauses ensure comprehensive coverage of activation scenarios. The description uses appropriate third-person voice and avoids vague language.

DimensionReasoningScore

Specificity

Lists multiple specific concrete actions: diagnosing failures, applying Anthropic best practices, and enumerates specific areas like explicit instructions, context/motivation, examples, output/verbosity control, thinking/effort, CLAUDE.md size, and skill-description rules.

3 / 3

Completeness

Clearly answers both 'what' (diagnose failures and apply Anthropic best practices for prompt engineering artifacts) and 'when' (explicit 'Use when' clause at the start plus a 'Triggers when' clause listing specific scenarios like prompts not followed, skills not activating, CLAUDE.md too long, or migrating prompts).

3 / 3

Trigger Term Quality

Excellent coverage of natural trigger terms users would say: 'CLAUDE.md', 'AGENTS.md', 'custom commands', 'skill files', 'prompt isn't followed', 'skill won't activate', 'CLAUDE.md is too long or ignored', 'migrating prompts'. These are highly natural phrases a user would use when encountering these specific problems.

3 / 3

Distinctiveness Conflict Risk

Highly distinctive niche focused specifically on optimizing Claude's configuration files (CLAUDE.md, AGENTS.md, skill files) using Anthropic best practices. The specific artifact types and failure scenarios make it very unlikely to conflict with other skills.

3 / 3

Total

12

/

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.

Validation11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository
joshuadavidthomas/agent-skills
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.