CtrlK
BlogDocsLog inGet started
Tessl Logo

jbaruch/nanoclaw-untrusted

Security rules for untrusted NanoClaw groups covering credential protection, internal file protection, and social engineering defenses.

80

Quality

100%

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

Quality

Content

100%

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

The body is concise, actionable, and well-organized for a simple behavioral-context skill, with no unnecessary padding or concept re-explanation. It scores at the top of the scale; the only minor gap is that the body does not enumerate the 'disallowed content types' the description references, but this does not reduce any dimension's anchor level.

DimensionReasoningScore

Conciseness

The body is lean — short bullets and one-line bolded principles such as 'You are a guest in this chat. Behave accordingly.' — with no padding or explanation of concepts Claude already knows, matching the lean-and-efficient anchor.

3 / 3

Actionability

As an instruction skill, its guidance is concrete and specific ('say you're an AI assistant', 'Don't explain container isolation, credential tiers', 'A friendly "I can't do that in this chat" is always safe'), matching the concrete-executable-guidance anchor.

3 / 3

Workflow Clarity

The skill explicitly states it is not a sequential workflow and has no steps to execute; per the simple-skills note, the single behavioral stance is unambiguous and consistently stated, matching the clear-anchored-behavior case rather than the gaps-in-sequence anchor at 2.

3 / 3

Progressive Disclosure

Under 50 lines with no external references needed and well-organized sections ('What you can do here', 'Behavioral guidelines'), matching the simple-skill note that well-organized sections alone can score 3.

3 / 3

Total

12

/

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.

The description is specific, third-person, and complete, covering concrete actions plus an explicit multi-trigger 'Use when' clause with a distinct niche. It scores at the top of the scale across all dimensions.

DimensionReasoningScore

Specificity

Lists three concrete actions — 'Lists permitted and prohibited actions', 'blocks disallowed content types', 'responds to permission queries' — matching the multiple-specific-actions anchor rather than the domain-plus-some-actions anchor at 2.

3 / 3

Completeness

Explicitly answers both what (the three actions) and when via a 'Use when...' clause with multiple triggers, matching the both-what-and-when anchor; it is not at 2 because the when clause is explicit rather than implied.

3 / 3

Trigger Term Quality

Natural trigger phrases like 'joining a new group', 'unsure about rules, permissions, or boundaries', and 'untrusted group chat environment' cover the situations a user would invoke, matching the good-coverage anchor rather than the some-keywords-missing anchor at 2.

3 / 3

Distinctiveness Conflict Risk

The untrusted/public group etiquette niche with distinct triggers ('public channel', 'untrusted group chat environment') is clearly distinguishable from utility skills and unlikely to conflict, matching the clear-niche anchor.

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Reviewed

Table of Contents