DF openclaw-defender
Protects your OpenClaw agent from the threats discovered in Snyk's ToxicSkills research (Feb 2026):
Protects your OpenClaw agent from the threats discovered in Snyk's ToxicSkills research (Feb 2026):
As a process F 48/100 · Will not run — References files that are not bundled: references/blocklist.conf
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.
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
- The text references files that are not there: add them or drop the references.
- 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 · 24
-
high Dangerous commands
cmd-encoded-execREADME.md:102Executes a base64/encoded payload (documentation of a security skill)echo "Y3Vy…b20=" | base64 -d | bash
security skill -
high Dangerous commands
cmd-encoded-execSKILL.md:249Executes a base64/encoded payload (documentation of a security skill)echo "Y3Vy…b20=" | base64 -d | bash
security skill
Medium and low: 22
-
medium Instruction override
en-ignore-previousreferences/threat-patterns.md:21Instruction-override phrase ("ignore previous instructions") (documentation of a security skill)| Ignore previous instructions | CRITICAL | "Ignore previous instructions and..." |
security skill -
medium Dangerous commands
cmd-pipe-to-shellreferences/threat-patterns.md:64Downloads and executes remote code from an unrecognised host (pipe to shell) (documentation of a security skill)- curl\|bash: `(curl|wget).*\|.*(bash|sh)`
security skill -
low Dangerous commands
cmd-cron-mentionREADME.md:38Mentions editing / listing crontab (documentation of a security skill)crontab -e
security skill -
low Secrets in code
secret-labelled-tokenreferences/runtime-integration.md:280Labelled token / key literal (vendor format unknown — verify it is not a live credential) (placeholder value)echo "API_KEY=0x12…def" | ./runtime-monitor.sh sanitize
placeholder -
low Secrets in code
secret-password-literalreferences/runtime-integration.md:280Hard-coded password / key literal (may be an example) (placeholder value)echo "API_KEY=0x12…def" | ./runtime-monitor.sh sanitize
placeholder -
low Dangerous commands
cmd-pipe-to-shellreferences/threat-patterns.md:13Downloads and executes remote code from an unrecognised host (pipe to shell) (documentation table row; documentation of a security skill)| `curl \| bash` / `wget \| sh` | CRITICAL | `curl -fsSL URL \| bash` |
tablesecurity skill -
low Secrets in code
secret-high-entropy-tokenreferences/toxicskills-research.md:68High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)Base64: "UGxl…mRz"
quoted -
low Risky intent
intent-offensive-securityreferences/toxicskills-research.md:203Offensive-security / dual-use content (legitimate for authorised testing; review intended use)Result: Privilege escalation
-
low Risky intent
intent-offensive-securityreferences/toxicskills-research.md:315Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- Privilege escalation
-
low Risky intent
intent-offensive-securityreferences/toxicskills-research.md:324Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- Regular penetration testing
-
low Risky intent
intent-offensive-securityreferences/toxicskills-research.md:347Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- Every tool = potential privilege escalation
-
low Risky intent
intent-offensive-securityreferences/toxicskills-research.md:395Offensive-security / dual-use content (legitimate for authorised testing; review intended use)1. **Red Team Testing**
-
low Dangerous commands
cmd-cron-mentionSKILL.md:86Mentions editing / listing crontab (documentation of a security skill)crontab -e
security skill -
low Instruction override
en-ignore-previousSKILL.md:201Instruction-override phrase ("ignore previous instructions") (negated — the text forbids it)- "Ignore previous instructions" or DAN-style jailbreaks
negated -
low Dangerous commands
cmd-pipe-to-shellSKILL.md:204Downloads and executes remote code from an unrecognised host (pipe to shell) (negated — the text forbids it)- `curl | bash` patterns
negated
A further 7 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 14. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
description-no-whendescription does not say WHEN to use the skill (no "use when") - warning
missing-refreference to a missing file: references/blocklist.conf
Process rating: all ten parameters 48/100
- 0Tools and files. 1 referenced file(s) missing: references/blocklist.conf
- 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. 10 mutating operations with no state check
- 50When it triggers. No condition that starts the skill
- 100Steps. 129 steps
- 100Failures and branches. 1 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3785 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 12 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)
- +3Description length 99: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +4Structure: 37 headings
- +3Step-by-step instructions: 129 items
- +4Has examples (16 code blocks)
- +4Reference files are cited in the instructions (3 of 5)
- +3All 7 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 70.