DC clawdefender
Security scanner and input sanitizer for AI agents. Detects prompt injection, command injection, SSRF, credential exfiltration, and path traversal attacks. Use when (1) installing new skills from ClawHub, (2) processing external input like emails, calendar events, Trello cards, or API responses, (3) validating URLs before fetching, (4) running security audits on your workspace. Protects agents from malicious content in untrusted data sources.
Security scanner and input sanitizer for AI agents.
As a process C 54/100 · Has gaps — weak spots: result and completion, when it triggers, failures and branches
What is at stake
The skill contains fragments that, in the wrong hands, cost money or data. Below: what the installer risks and what the author should do.
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.
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
- Address the high-severity findings: each costs 18 safety points. If one is a false positive, add the rule id to guard.allow in spec.yaml.
- 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 · 12
-
high Dangerous commands
cmd-encoded-execscripts/clawdefender.sh:160Executes a base64/encoded payload (string literal in code, not executed)'base64 -d \| bash'
code literal -
high Dangerous commands
cmd-encoded-execscripts/clawdefender.sh:161Executes a base64/encoded payload (string literal in code, not executed)'base64 --decode \| sh'
code literal
Medium and low: 10
-
medium Instruction override
en-ignore-previousscripts/clawdefender.sh:74Instruction-override phrase ("ignore previous instructions") (quoted — discussed, not commanded)'ignore previous instructions'
quoted -
medium Dangerous commands
cmd-pipe-to-shellscripts/clawdefender.sh:158Downloads and executes remote code from an unrecognised host (pipe to shell) (string literal in code, not executed)'curl.*\| bash'
code literal -
medium Dangerous commands
cmd-pipe-to-shellscripts/clawdefender.sh:159Downloads and executes remote code from an unrecognised host (pipe to shell) (string literal in code, not executed)'wget.*\| sh'
code literal -
low Dangerous commands
cmd-privilegescripts/clawdefender.sh:147Privilege escalation / world-writable permissions (string literal in code, not executed)'chmod 777'
code literal -
low Dangerous commands
cmd-destructive-fsSKILL.md:162Destructive filesystem command (wipes root/home/drive) (negated — the text forbids it)- Fork bombs `:(){ :|:& };:`negated -
low Dangerous commands
cmd-privilegeSKILL.md:164Privilege escalation / world-writable permissions (negated — the text forbids it)- `chmod 777`, `eval`, `exec`
negated
A further 4 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
✓ No remarks against the Agent Skills spec
Process rating: all ten parameters 54/100
- 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
- 60Tools and files. Uses tools (bash, web) that frontmatter does not declare
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Steps. 33 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1590 tokens
- 100Running it twice. No mutating operations
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 446: enough signal without eating the budget
- +4Structure: 24 headings
- +3Step-by-step instructions: 33 items
- +4Has examples (14 code blocks)
- +3All 2 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 87.