CtrlK
BlogDocsLog inGet started
Tessl Logo

self-customize

Customize your own agent — add capabilities, install packages, add MCP servers, edit code or CLAUDE.md. Use when the user asks you to add a feature, install a tool, or modify how you work. For non-trivial code changes, delegate to a builder agent via create_agent.

67

Quality

80%

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 ./container/skills/self-customize/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

77%Weight 40%Scale 1-3

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

A highly actionable, well-sequenced skill body with concrete tool calls, a reusable builder template, and explicit validation/feedback loops. Its weak spots are the explanatory 'Why' rationale that pads the token budget and the lack of progressive disclosure via reference files.

Suggestions

Trim the 'Diff Size Limits — Why' section to the policy and the one-line decomposition rationale; Claude already understands that smaller diffs are more reviewable.

Extract the builder agent CLAUDE.md template into a references/builder-instructions.md file and link to it one level deep, so the main SKILL.md stays a lean overview.

DimensionReasoningScore

Conciseness

The body is mostly efficient and actionable, but the 'Diff Size Limits — Why' section explains reviewability Claude already knows ('A 50-line focused change is reviewable. A 500-line sweep is not.', 'Makes human approval meaningful (you can actually read 150 lines)') and could be trimmed.

2 / 3

Actionability

Fully executable, copy-paste-ready guidance: concrete tool calls with arguments (install_packages({ apt: [...], npm: [...], reason: '...' }), add_mcp_server({ name, command, args }), create_agent({ name, instructions })) plus a complete builder CLAUDE.md template and two worked examples.

3 / 3

Workflow Clarity

Clear numbered sequences with explicit validation and feedback checkpoints: the admin-approval gate, the builder's report-back step, 'Test if possible', 'If a change would break existing tests, stop and report', and the main agent's review of the builder summary.

3 / 3

Progressive Disclosure

Sections are well organized with no nested references, but the skill is a monolithic ~83-line file (over the 50-line simple-skill threshold) with an embedded ~18-line builder CLAUDE.md template that could live in a separate reference file; no one-level-deep file references are signaled.

2 / 3

Total

10

/

12

Passed

Description

82%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A well-constructed description that pairs concrete customization actions with an explicit 'Use when' trigger clause and natural user vocabulary. Its main weaknesses are second-person voice (which caps specificity) and somewhat broad 'add a feature'/'install a tool' triggers that raise conflict risk.

Suggestions

Rewrite in third person to avoid the specificity penalty, e.g. 'Customizes the agent — adds capabilities, installs packages, adds MCP servers, edits code or CLAUDE.md. Use when the user requests a new feature, tool install, or workflow change.'

Tighten the trigger terms to reduce conflict with general dev skills, e.g. lead with 'customize', 'self-modify', or 'add an MCP server' rather than the generic 'add a feature'.

DimensionReasoningScore

Specificity

It lists multiple concrete actions ('add capabilities, install packages, add MCP servers, edit code or CLAUDE.md', 'delegate to a builder agent via create_agent') that would merit a 3, but the second-person phrasing ('your own agent', 'the user asks you to add', 'modify how you work') triggers the rubric's -1 specificity penalty.

2 / 3

Completeness

Clearly answers both 'what' (the enumerated customization mechanisms) and 'when' via an explicit 'Use when the user asks you to add a feature, install a tool, or modify how you work' clause.

3 / 3

Trigger Term Quality

Strong coverage of natural terms a user would actually say ('add a feature', 'install a tool', 'install packages', 'modify how you work'), matching the 'good coverage of natural terms' anchor.

3 / 3

Distinctiveness Conflict Risk

The 'customize your own agent' framing with MCP/packages/CLAUDE.md mechanisms is a niche, but the 'add a feature' and 'install a tool' triggers are broad and could overlap with general development skills, matching the 'somewhat specific but could still overlap' anchor.

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
nanocoai/nanoclaw
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.