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

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/context-engineering/SKILL.md

Quality

Content

71%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 highly actionable with excellent copy-paste templates and clear sub-workflows, but it is a long monolithic file with no progressive disclosure offloading and notable redundancy across its diagnostic sections. Tightening and splitting reference material would materially improve token efficiency.

Suggestions

Move the full CLAUDE.md template, the MCP server table, and the other-tool rules-file mapping into a references/ file, keeping SKILL.md as a concise overview with clearly signaled one-level-deep links.

Consolidate the overlapping "Red Flags", "Common Rationalizations", and "Anti-Patterns" sections into a single table; each anti-pattern already restates a red flag and its fix.

Trim explanatory prose Claude already knows (e.g., the lost-in-the-middle citation explanation and the "context is the single biggest lever" framing) down to the actionable rule.

Dimension	Reasoning	Score
Conciseness	Mostly prescriptive and example-driven, but ~350 lines include unnecessary material: framing prose ("Context is the single biggest lever"), explanation of known effects (lost-in-the-middle), and heavy redundancy between "Red Flags", "Common Rationalizations", "Anti-Patterns", and "When to Use". Not 2: the bulk is tight tables and templates; not 4: three overlapping diagnostic sections is more than minor trimming.	3 / 5
Actionability	Fully executable, copy-paste-ready artifacts throughout: a complete CLAUDE.md template, Selective Include and Project Map blocks, CONFUSION/MISSING REQUIREMENT surfacing templates, the inline PLAN pattern, and a before/after compression example. Not 4: these are concrete usable templates covering the common cases, not high-level hints.	5 / 5
Workflow Clarity	Clear numbered sequences (pre-task context loading 1-4, restartable-session boundaries with re-verify conditions) plus a final Verification checklist provide most checkpoints. Not 5: the skill is a practice guide with several parallel sub-workflows rather than one end-to-end sequence with explicit validate-fix-retry feedback loops; not 3: checkpoints are explicit and present in each procedure.	4 / 5
Progressive Disclosure	The single 350-line file has well-organized sections and headers, but no bundle files exist and detailed material that belongs in separate references (full CLAUDE.md template, MCP server table, other-tool rules-file mapping) is inlined. Not 4: nothing is offloaded to reference files despite the length.	3 / 5
	Total	15 / 20 Passed

Description

70%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 has a strong, explicit trigger section with four natural use-when phrases, but the capability statement is a single generic action that undersells the skill's concrete scope. Distinctiveness is good with minor overlap risk against configuration-adjacent skills.

Suggestions

Replace the generic "Optimizes agent context setup" with 2-3 concrete actions, e.g. "Creates rules files (CLAUDE.md), curates per-task context, and manages context budget during long sessions."

Add natural trigger synonyms users actually say, such as "hallucinating APIs", "ignoring conventions", or "CLAUDE.md", to the Use when clause.

Dimension	Reasoning	Score
Specificity	"Optimizes agent context setup" names the domain and, with "configure rules files and context", gives two explicit actions, matching the anchor for 1-2 concrete but non-comprehensive actions. Not 4: the what-clause is a single generic verb and omits much of what the skill body covers (specs, context budgeting, session boundaries).	3 / 5
Completeness	Both are present: an explicit "Use when..." clause with four concrete triggers plus a what-clause. Not 5: the what ("Optimizes agent context setup") is thin and doesn't concretely describe what the skill does; not 3 because the when-guidance is explicit and specific.	4 / 5
Trigger Term Quality	"starting a new session", "agent output quality degrades", "switching between tasks", "configure rules files" are natural phrases a user would say, giving good keyword coverage. Not 5: common synonyms like CLAUDE.md, hallucinating, or context window are missing.	4 / 5
Distinctiveness Conflict Risk	"agent context setup" carves a mostly distinct niche with trigger phrases unlikely to fire for unrelated skills. Not 5: some overlap risk with memory-management or editor-config skills that also mention rules files.	4 / 5
	Total	15 / 20 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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: addyosmani/agent-skills
Path: skills/context-engineering/SKILL.md
Commit: 2686b62

Reviewed: 3 days 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.