CtrlK
BlogDocsLog inGet started
Tessl Logo

defense-in-depth-validation

Validate at every layer data passes through to make bugs impossible. Use when invalid data causes failures deep in execution, requiring validation at multiple system layers.

81

1.42x
Quality

72%

Does it follow best practices?

Impact

100%

1.42x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./plugins/defense-in-depth-validation/skills/defense-in-depth-validation/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

100%

60%

Project Creation Service Hardening

Four-layer input validation

Criteria
Without context
With context

Entry: empty check

100%

100%

Entry: existence check

0%

100%

Entry: directory type check

0%

100%

Business logic guard

100%

100%

Environment guard: test detection

0%

100%

Environment guard: path boundary

0%

100%

Debug: stack trace captured

0%

100%

Debug: context fields logged

0%

100%

Distinct error messages

100%

100%

Validation plan documents layers

100%

100%

100%

10%

CI Workspace Cleanup Tool Safety

Environment guards and debug instrumentation

Criteria
Without context
With context

Test env detection

100%

100%

tmpdir boundary check

100%

100%

Refusal outside tmpdir

100%

100%

Error message identifies context

100%

100%

Stack trace captured

100%

100%

Directory logged

100%

100%

cwd logged

0%

100%

Debug before destructive op

100%

100%

Path normalization

100%

100%

Safety notes cover both mechanisms

100%

100%

100%

20%

Session Manager Bug: Tracing and Fixing Invalid Data Propagation

Data flow tracing and layer-bypass testing

Criteria
Without context
With context

Entry validation added

100%

100%

Business logic validation added

100%

100%

Distinct error messages per layer

0%

100%

Layer bypass test present

100%

100%

Entry layer test present

100%

100%

Environment guard or debug instrumentation

0%

100%

Data flow documents origin

100%

100%

Data flow documents checkpoints

100%

100%

No single-point fix

100%

100%

Tests are independent

100%

100%

Repository
secondsky/claude-skills
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

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.