context-engineering

Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and context for a project.

Quality

42%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/context-engineering/SKILL.md

Quality

Discovery

57%

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 has a strong 'Use when...' clause with multiple explicit trigger scenarios, which is its main strength. However, it suffers from vague capability description—'optimizes agent context setup' is abstract and doesn't tell Claude what concrete actions the skill performs. The trigger terms are decent but could be more specific to the actual tools and file types involved.

Suggestions

Replace 'Optimizes agent context setup' with specific concrete actions, e.g., 'Creates and configures rules files (.clinerules, CLAUDE.md, .cursorrules), generates system prompts, and structures project context for AI coding agents.'

Add more natural trigger terms that users would actually say, such as specific file names, tool names (Cursor, Cline, Claude Code), or phrases like 'set up my agent', 'improve agent performance', 'write a rules file'.

Clarify the domain more precisely to reduce overlap—specify which agents or frameworks this applies to, distinguishing it from general project configuration or session management skills.

Dimension	Reasoning	Score
Specificity	The description says 'Optimizes agent context setup' which is vague and abstract. It does not list any concrete actions like 'creates rules files', 'configures .cursorrules', or 'generates system prompts'. The phrase 'configure rules files and context' is slightly more specific but still lacks concrete detail about what actions are performed.	1 / 3
Completeness	The description answers both 'what' (optimizes agent context setup, configure rules files and context) and 'when' with explicit triggers ('starting a new session', 'output quality degrades', 'switching between tasks', 'need to configure rules files'). The 'Use when...' clause is present and detailed.	3 / 3
Trigger Term Quality	Includes some relevant terms like 'session', 'rules files', 'context', 'output quality degrades', and 'switching between tasks'. However, it misses natural user phrases like 'system prompt', 'agent instructions', '.clinerules', 'CLAUDE.md', or specific tool/framework names that would help match user requests.	2 / 3
Distinctiveness Conflict Risk	The term 'agent context setup' provides some specificity, but phrases like 'starting a new session' and 'switching between tasks' are quite broad and could overlap with general project setup, onboarding, or task management skills. The mention of 'rules files' adds some distinctiveness but the overall scope is still somewhat ambiguous.	2 / 3
	Total	8 / 12 Passed

Implementation

27%

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

This skill reads more like a comprehensive blog post or tutorial on context engineering than a concise, actionable skill file for Claude. Its main strengths are the concrete CLAUDE.md template and the structured context hierarchy, but these are buried in extensive philosophical discussion, rationalizations tables, and explanations of concepts Claude already understands. The content would benefit enormously from being cut by 60-70% and split across multiple files.

Suggestions

Cut the content to under 80 lines by removing sections Claude already knows (overview explanation, common rationalizations, red flags that restate anti-patterns) and keeping only the CLAUDE.md template, context hierarchy diagram, and pre-task loading checklist.

Split detailed content into referenced files: e.g., CONTEXT_PACKING.md for the three packing strategies, CONFUSION_MANAGEMENT.md for ambiguity handling patterns, and MCP_INTEGRATIONS.md for the MCP server table.

Add explicit validation steps to the workflow: after setting up context, include a concrete verification step like 'Ask the agent to describe the project conventions back to you' before proceeding with tasks.

Remove the 'Common Rationalizations' and 'Confusion Management' sections entirely — these teach Claude things it already knows about being a good assistant and handling ambiguity.

Dimension	Reasoning	Score
Conciseness	This skill is extremely verbose at ~250+ lines, with significant content Claude already knows (what context engineering is, why context matters, how to manage conversations). Sections like 'Common Rationalizations' and 'Confusion Management' explain general agent interaction principles that don't add actionable value. The overview paragraph explains concepts Claude inherently understands.	1 / 3
Actionability	The skill provides concrete examples like the CLAUDE.md template, context packing strategies with specific formats, and the anti-patterns table. However, much of the content is advisory/philosophical rather than executable — there are no actual commands to run, no scripts to execute, and the guidance is more about principles than specific actions Claude should take in a given situation.	2 / 3
Workflow Clarity	There's a clear hierarchy (Levels 1-5) and some sequencing (pre-task context loading steps, inline planning pattern), but the skill lacks explicit validation checkpoints. The verification checklist at the end is good but disconnected from the workflow. There's no clear 'do this, then verify, then proceed' feedback loop for the context setup process itself.	2 / 3
Progressive Disclosure	This is a monolithic wall of text with no references to supporting files. At this length (~250+ lines), content like the MCP integrations table, confusion management patterns, context packing strategies, and anti-patterns tables should be split into separate referenced documents. Everything is inlined in a single massive file with no bundle files to support it.	1 / 3
	Total	6 / 12 Passed

Validation

100%

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 — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: addyosmani/agent-skills
Commit: f17c6e8

Reviewed: about 11 hours 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.