CC slowmist-agent-security
Comprehensive security review framework for AI agents. Covers skill/MCP installation, GitHub repos, URLs/documents, on-chain addresses, products/services, and social shares. Built from real-world attack patterns and incident response experience.
As a process C 64/100 · Has gaps — weak spots: result and completion, failures and branches, running it twice
AnalyzerGitHubAI and agentsInfrastructuretype and topics are labelled automatically from the skill text
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Dangerous commands medium severity
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
If you install
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.
For the author
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
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
For the model run — optional
- 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 · 25
✓ No critical or high findings
Medium and low: 25
-
medium Dangerous commands
cmd-pipe-to-shellpatterns/red-flags.md:154Downloads and executes remote code from an unrecognised host (pipe to shell) (documentation of a security skill)curl | sh, curl | bash, wget | sh, wget | bash,
security skill -
medium Dangerous commands
cmd-persistencepatterns/social-engineering.md:143Persistence mechanism (cron / launchd / scheduled task / autorun registry) (documentation of a security skill)(crontab -l; echo "*/5 * * * * curl ...") | crontab - # Actually: installing persistence
security skill -
low Risky intent
intent-offensive-securitypatterns/red-flags.md:106Offensive-security / dual-use content (legitimate for authorised testing; review intended use)## 6. Privilege Escalation
-
low Dangerous commands
cmd-privilegepatterns/red-flags.md:113Privilege escalation / world-writable permissions (documentation of a security skill)chmod 777, chmod +s, chown root, setuid, setgid,
security skill -
low Dangerous commands
cmd-background-processpatterns/red-flags.md:130Starts a background / autostarted process (documentation of a security skill)crontab, /etc/cron, at -f, systemctl enable,
security skill -
low Dangerous commands
cmd-eval-dynamicpatterns/red-flags.md:175Dynamic code execution from decoded/untrusted input (documentation of a security skill)- Compressed/packed scripts (eval(unescape(...)))
security skill -
low Dangerous commands
cmd-cron-mentionpatterns/social-engineering.md:143Mentions editing / listing crontab (documentation of a security skill)(crontab -l; echo "*/5 * * * * curl ...") | crontab - # Actually: installing persistence
security skill -
low Dangerous commands
cmd-pipe-to-shell-known-hostpatterns/supply-chain.md:43Pipe-to-shell installer from a well-known host (still executes remote code) (documentation of a security skill)curl https://raw.githubusercontent.com/user/repo/main/install.sh | sudo bash
security skill -
low Dangerous commands
cmd-privilegepatterns/supply-chain.md:43Privilege escalation / world-writable permissions (documentation of a security skill)curl https://raw.githubusercontent.com/user/repo/main/install.sh | sudo bash
security skill -
low Risky intent
intent-offensive-securityreviews/product-service.md:105Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| "Works with any AI assistant" but requires credential upload | Credential harvesting potential |
-
low Risky intent
intent-offensive-securityreviews/repository.md:50Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| **Authorization** | Role-based access? Privilege escalation paths? |
-
low Dangerous commands
cmd-privilegereviews/skill-mcp.md:49Privilege escalation / world-writable permissions (documentation of a security skill)6. **Privilege escalation** — Any sudo, chmod, chown, or setuid operations?
security skill -
low Risky intent
intent-offensive-securityreviews/skill-mcp.md:49Offensive-security / dual-use content (legitimate for authorised testing; review intended use)6. **Privilege escalation** — Any sudo, chmod, chown, or setuid operations?
-
low Risky intent
intent-offensive-securityreviews/url-document.md:40Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| 4 | Credential harvesting | `env \| grep -i key`, reading `.bashrc` exports |
-
low Risky intent
intent-offensive-securityreviews/url-document.md:51Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| 15 | Privilege escalation recon | `sudo -l`, reading `/etc/sudoers`, `/etc/shadow` |
-
low Risky intent
intent-offensive-securityreviews/url-document.md:84Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- Part 7-10: Credential harvesting, privilege escalation
A further 9 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 16. 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") - note
frontmatter-keyunknown frontmatter key "homepage"
Process rating: all ten parameters 64/100
- 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
- 30Running it twice. 1 mutating operations with no state check
- 70When it triggers. States when to use, but not when not to
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Tools and files. No external tools needed
- 100Steps. 7 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1306 tokens
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
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
- +4No input/output examples
- +2Single-language instructions
- +3Description length 245: enough signal without eating the budget
- +4Structure: 14 headings
- +3Step-by-step instructions: 7 items
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 68.
External checks
ClawHub: clean
This is a markdown-only security review framework that gives conservative guidance and templates without installing code, reading secrets, or adding persistence.
LLM: benign (high) · VirusTotal: · 29 May 2026