CtrlK
BlogDocsLog inGet started
Tessl Logo

project-structure

Mandatory practices for reading, understanding, and maintaining project documentation (CLAUDE.md, ARCHITECTURE.md). This skill MUST be used for ALL coding tasks, ALL conversations, ALL projects, regardless of language or context. It defines non-negotiable rules for how to orient yourself in any codebase before doing any work. Trigger on every single interaction — any file edit, code review, bug fix, feature addition, refactoring, question about code, or any task that involves a project directory. There are no exceptions. Even if the user asks a simple question, these rules apply.

53

Quality

58%

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 ./.claude/skills/project-structure/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

57%

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

The body is well-organized and gives genuinely specific behavioral rules, but it is padded with emphatic filler and lacks concrete mechanisms or validation checkpoints for verifying documentation coverage. As an instruction-only skill it is actionable enough, yet could be tightened and made verifiable.

Suggestions

Remove the redundant emphatic filler ('No matter what. No matter how small the task seems. No matter if you think you already know the answer. Read them every time.') and state the rule once.

Add a concrete verification step for the coverage rule, e.g. 'After editing, confirm a CLAUDE.md exists next to each Cargo.toml/package.json in the changed subtree.'

Provide a concrete mechanism for the hierarchy read (e.g., 'walk parent directories from cwd to root, reading each CLAUDE.md') instead of only describing it abstractly.

DimensionReasoningScore

Conciseness

The body is reasonably lean and mostly instructions rather than concept explanation, but it pads with emphatic filler ('NEVER skip reading these files. No matter what. No matter how small the task seems. No matter if you think you already know the answer. Read them every time.') that adds tokens without adding information.

2 / 3

Actionability

It gives concrete, specific guidance (read CLAUDE.md up the hierarchy, create a sibling CLAUDE.md for every Cargo.toml/package.json), but offers no executable commands or mechanism for locating files, so it describes a procedure rather than giving copy-ready steps.

2 / 3

Workflow Clarity

Steps are sequenced (read hierarchy bottom-up, read sibling ARCHITECTURE.md, update on learnings, ensure coverage), but there are no validation checkpoints — nothing tells Claude how to confirm it has read the right files or verify coverage is complete, so checkpoints are only implicit.

2 / 3

Progressive Disclosure

For a sub-50-line single-purpose skill with no bundle files, the content is well-organized into clearly labeled sections ('Reading Project Documentation', 'ARCHITECTURE.md', 'Maintaining Project Documentation', 'Ensuring CLAUDE.md Coverage'), which meets the simple-skill bar for a 3.

3 / 3

Total

9

/

12

Passed

Description

60%

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 clearly states what the skill does and when to trigger it, but severely over-claims by mandating use for every possible interaction, which destroys distinctiveness and inflates conflict risk. Trigger terms are present but framed as blanket imperatives rather than natural user phrasing.

Suggestions

Narrow the trigger to genuine documentation-orientation moments (e.g., 'Use when entering an unfamiliar codebase, before edits to a new module, or when project conventions matter') instead of claiming it applies to ALL tasks with no exceptions.

Rewrite the imperative 'MUST be used for ALL coding tasks' framing into third-person natural phrasing a user would actually say, e.g., 'Orients to a project by reading CLAUDE.md/ARCHITECTURE.md before working.'

Drop the redundant blanket statements ('There are no exceptions', 'Even if the user asks a simple question') that pad the description without adding trigger specificity.

DimensionReasoningScore

Specificity

It names concrete actions ('reading, understanding, and maintaining project documentation', 'read all CLAUDE.md files', 'Update CLAUDE.md IMMEDIATELY'), but the verbs are generic orientation/maintenance tasks rather than the multiple distinct concrete operations that anchor a 3.

2 / 3

Completeness

It explicitly answers both 'what' ('Mandatory practices for reading, understanding, and maintaining project documentation') and 'when' ('Trigger on every single interaction — any file edit, code review, bug fix...'), satisfying the explicit-trigger requirement for a 3.

3 / 3

Trigger Term Quality

It lists many task types ('file edit, code review, bug fix, feature addition, refactoring, question about code') and named files ('CLAUDE.md', 'ARCHITECTURE.md'), which are relevant triggers, but it over-relies on imperative framing ('MUST be used for ALL coding tasks') rather than the natural phrases a user would actually say.

2 / 3

Distinctiveness Conflict Risk

The description claims applicability to 'ALL coding tasks, ALL conversations, ALL projects' with 'no exceptions', which makes it maximally generic and guaranteed to conflict with essentially every other skill rather than occupying a clear niche.

1 / 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
enso-org/enso
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.