AB skill-vetter-v2
Verification-guided review workflow for inspecting skill packages before use or publication. Classifies risk and flags claims that exceed evidence. Does not itself perform cryptographic verification, emit receipts, or prove a skill is safe.
As a process B 65/100 · Nearly there — weak spots: when it triggers, running it twice, progress reporting
How to improve
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 3
✓ No critical or high findings
Medium and low: 3
-
low Risky intent
intent-offensive-securityassets/REVIEW-CHECKLIST.md:23Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- privilege escalation
-
low Dangerous commands
cmd-privilegescripts/scan-skill.sh:37Privilege escalation / world-writable permissions (string literal in code, not executed; documentation of a security skill)[privilege]='sudo|chmod 777|chown '
code literalsecurity skill -
low Risky intent
intent-offensive-securitySKILL.md:146Offensive-security / dual-use content (legitimate for authorised testing; review intended use)* privilege escalation or system-level modification
Files scanned: 13. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
✓ No remarks against the Agent Skills spec
Process rating: all ten parameters 65/100
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 3 mutating operations with no state check
- 55Failures and branches. 1 branches
- 60Tools and files. Uses tools (bash) that frontmatter does not declare
- 60Result and completion. Output format stated, no completion criterion
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Steps. 74 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 2592 tokens
- low 16 top-level sections: this looks like several domains in one skill
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- -42 reference files, but SKILL.md never points to them: the model will not open them
- -31 of 2 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +3Description length 240: enough signal without eating the budget
- +4Structure: 25 headings
- +3Step-by-step instructions: 74 items
- +3Output format is stated explicitly
- +4Has examples (6 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 80.