BC security-auditor
Autonomously scans all installed OpenClaw skills for security risks. Detects dangerous behaviors like shell execution, file deletion, remote code download, data exfiltration, and obfuscated logic. Assigns each skill a risk score (0–100), risk level (Low/Medium/High), and generates a full security report with mitigation recommendations. Use this when the user asks to audit skills, check for security risks, scan installed skills, or wants a security report.
As a process C 54/100 · Has gaps — weak spots: result and completion, inputs and preconditions, consistency
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 skill contains commands that delete files, rewrite disks or execute code fetched from the network. The agent may run them without asking if it believes the instructions require it.
Replace destructive commands with safe equivalents that ask for confirmation, scope them to one folder, and stop piping curl into a shell: pin a version and a checksum.
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 · 19
✓ No critical or high findings
Medium and low: 19
-
low Concealment
en-hide-from-userdata/example-output.md:60Instruction to hide actions from the user (test fixture / example file; documentation of a security skill)files in ~/Documents or ~/Desktop and silently upload them to an external
fixturesecurity skill -
low Risky intent
intent-offensive-securityREADME.md:206Offensive-security / dual-use content (legitimate for authorised testing; review intended use)High risk (H1–H16): shell execution, remote code download, file deletion, obfuscation, privilege escalation, credential harvesting, .env access, keyloggers, clipboard theft, screen capture, crypto min
-
low Risky intent
intent-offensive-securitySKILL.md:105Offensive-security / dual-use content (legitimate for authorised testing; review intended use)**H5 — Privilege escalation**
-
low Risky intent
intent-offensive-securitySKILL.md:107Offensive-security / dual-use content (legitimate for authorised testing; review intended use) (detector / deny-list definition)Finding: "Attempts privilege escalation — can gain elevated OS permissions."
detector -
low Risky intent
intent-offensive-securitySKILL.md:119Offensive-security / dual-use content (legitimate for authorised testing; review intended use)**H8 — Keylogger / input capture**
-
low Risky intent
intent-offensive-securitySKILL.md:121Offensive-security / dual-use content (legitimate for authorised testing; review intended use) (quoted — discussed, not commanded)Finding: "Captures keyboard input — potential keylogger, passwords and input silently recorded."
quoted -
low Risky intent
intent-offensive-securitySKILL.md:135Offensive-security / dual-use content (legitimate for authorised testing; review intended use)**H12 — Reverse shell / backdoor**
A further 12 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 10. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- note
frontmatter-keyunknown frontmatter key "runtime" - note
frontmatter-keyunknown frontmatter key "install" - note
frontmatter-keyunknown frontmatter key "requires" - note
frontmatter-keyunknown frontmatter key "permissions"
Process rating: all ten parameters 54/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 30Running it twice. 15 mutating operations with no state check
- 40Consistency. Frontmatter name (security-auditor) differs from the folder (claw-security-auditor)
- 60Tools and files. Uses tools (bash, web, node) that frontmatter does not declare
- 70When it triggers. States when to use, but not when not to
- 70Failures and branches. 5 branches
- 100Steps. 29 steps
- 100Execution cost. Instruction body is 3995 tokens
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 11 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
- -5TODO / placeholder text left in the skill
- -32 of 5 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +3Description length 459: enough signal without eating the budget
- +4Structure: 18 headings
- +3Step-by-step instructions: 29 items
- +4Has examples (3 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 73.