BD security-audit
Audit codebases, infrastructure, AND agentic AI systems for security issues. Covers traditional security (dependencies, secrets, OWASP web top 10, SSL/TLS, file permissions) PLUS agentic security (prompt injection scanning, identity spoofing detection, memory poisoning checks, multi-agent communication audit, OWASP Agentic Top 10). Use when scanning for vulnerabilities, detecting hardcoded secrets, reviewing agent workspace configuration, checking prompt injection vectors, or auditing agent permissions and boundaries.
Audit codebases, infrastructure, AND agentic AI systems for security issues.
As a process D 42/100 · Unfinished process — weak spots: result and completion, when it triggers, failures and branches
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The text contains phrases like "ignore previous instructions" or "you are now…". That is an attempt to hijack the agent: it may break your rules, the system limits or company policy.
An honest skill does not need them: state the role and the rules directly without overriding other instructions. Otherwise catalog scanners and corporate filters will block the listing.
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- 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 · 6
✓ No critical or high findings
Medium and low: 6
-
medium Instruction override
en-ignore-previousSKILL.md:623Instruction-override phrase ("ignore previous instructions") (documentation of a security skill)# Check for instructions in memory that override safety rules
security skill -
low Secrets in code
secret-password-literalSKILL.md:446Hard-coded password / key literal (may be an example) (placeholder value)API_KEY="sk-a…..."
placeholder -
low Secrets in code
secret-password-literalSKILL.md:453Hard-coded password / key literal (may be an example) (placeholder value)API_KEY=sk-a…...
placeholder -
low Dangerous commands
cmd-cron-mentionSKILL.md:674Mentions editing / listing crontab (documentation of a security skill)crontab -l 2>/dev/null || echo " No crontab"
security skill
A further 2 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 3. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 7525 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 42/100
- 0Result and completion. Does not say what the result is
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 2 mutating operations with no state check
- 40Consistency. Frontmatter name (security-audit) differs from the folder (agentic-security-audit)
- 60Tools and files. Uses tools (bash, python) that frontmatter does not declare
- 70Inputs and preconditions. Inputs and preconditions are listed
- 70Execution cost. Instruction body is 7525 tokens
- 85Steps. 23 steps, 1 vague phrases
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 10 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)
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +3Description length 523: enough signal without eating the budget
- +4Structure: 42 headings
- +3Step-by-step instructions: 23 items
- +4Has examples (31 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 74.