DD agentguard
GoPlus AgentGuard — AI agent security guard. Run /agentguard checkup for a full security health check, scans all installed skills, checks credentials, permissions, and network exposure, then delivers an HTML report directly to you. Also use for scanning third-party code, blocking dangerous commands, preventing data leaks, evaluating action safety, and running daily security patrols.
GoPlus AgentGuard — AI agent security guard.
As a process D 44/100 · Unfinished process — weak spots: result and completion, when it triggers, inputs and preconditions
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 purpose itself is risky: wallets, browser password stores, offensive security. Even an honest implementation gives the agent access to things that cost money.
Explain in the description why the access is needed and how it is limited; add tests that show refusals on dangerous requests.
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.
- 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 · 32
-
high Dangerous commands
cmd-destructive-fsSKILL.md:603Destructive filesystem command (wipes root/home/drive) (documentation of a security skill)| 2025-01-15 14:30 | Bash | rm -rf / | DENY | critical | DANGEROUS_COMMAND | some-skill |
security skill
Medium and low: 31
-
medium Dangerous commands
cmd-destructive-fsaction-policies.md:65Destructive filesystem command (wipes root/home/drive) (documentation table row; documentation of a security skill)| `> /dev/sda` | Disk overwrite |
tablesecurity skill -
medium Dangerous commands
cmd-persistencepatrol-checks.md:176Persistence mechanism (cron / launchd / scheduled task / autorun registry) (documentation of a security skill)ls -la /etc/cron.d/ /etc/cron.daily/ /etc/cron.hourly/ 2>/dev/null
security skill -
medium Dangerous commands
cmd-encoded-execpatrol-checks.md:195Executes a base64/encoded payload (documentation table row; documentation of a security skill)| `base64 -d \| bash` | Decode and execute | CRITICAL |
tablesecurity skill -
medium Dangerous commands
cmd-persistencepatrol-checks.md:219Persistence mechanism (cron / launchd / scheduled task / autorun registry) (documentation of a security skill)find /etc/cron.d/ -type f -mtime -1 2>/dev/null
security skill -
medium Risky intent
intent-wallet-secretsscan-rules.md:115Handles crypto-wallet secrets (seed / mnemonic / private key) — a classic stealer target| Quoted string with 12-24 BIP-39 words starting with: abandon, ability, able, about, above, absent, absorb, abstract, absurd, abuse | Mnemonic phrase |
-
low Dangerous commands
cmd-pipe-to-shellaction-policies.md:67Downloads and executes remote code from an unrecognised host (pipe to shell) (documentation table row; documentation of a security skill)| `wget\|sh` / `curl\|sh` | Download and execute |
tablesecurity skill -
low Dangerous commands
cmd-pipe-to-shellaction-policies.md:68Downloads and executes remote code from an unrecognised host (pipe to shell) (documentation table row; documentation of a security skill)| `wget\|bash` / `curl\|bash` | Download and execute |
tablesecurity skill -
low Exfiltration
exfil-read-secret-filesaction-policies.md:75Reads credential / secret files (documentation table row; documentation of a security skill)| `cat /etc/shadow` | Password hashes |
tablesecurity skill -
low Exfiltration
exfil-read-secret-filesaction-policies.md:79Reads credential / secret files (documentation table row; documentation of a security skill)| `cat ~/.npmrc` | npm auth tokens |
tablesecurity skill -
low Exfiltration
exfil-read-secret-filesaction-policies.md:80Reads credential / secret files (documentation table row; documentation of a security skill)| `cat ~/.netrc` | Network credentials |
tablesecurity skill -
low Dangerous commands
cmd-cron-mentionpatrol-checks.md:173Mentions editing / listing crontab (documentation of a security skill)crontab -l 2>/dev/null
security skill -
low Dangerous commands
cmd-eval-dynamicscan-rules.md:48Dynamic code execution from decoded/untrusted input (documentation of a security skill)| `exec\s*\(\s*requests\.get` | Python exec(requests.get()) |
security skill -
low Dangerous commands
cmd-eval-dynamicscan-rules.md:49Dynamic code execution from decoded/untrusted input (documentation of a security skill)| `eval\s*\(\s*requests\.get` | Python eval(requests.get()) |
security skill -
low Dangerous commands
cmd-eval-dynamicscan-rules.md:50Dynamic code execution from decoded/untrusted input (documentation of a security skill)| `exec\s*\(\s*urllib` | Python exec(urllib) |
security skill -
low Dangerous commands
cmd-eval-dynamicscan-rules.md:51Dynamic code execution from decoded/untrusted input (documentation of a security skill)| `eval\s*\(\s*urllib` | Python eval(urllib) |
security skill -
low Exfiltration
intent-browser-credential-storescan-rules.md:93Accesses a browser credential / cookie store (documentation table row; documentation of a security skill)| `Chrome.*Cookies` (i) | Chrome cookies |
tablesecurity skill -
low Risky intent
intent-wallet-secretsscan-rules.md:116Handles crypto-wallet secrets (seed / mnemonic / private key) — a classic stealer target (detector / deny-list definition; documentation table row)| `seed[_\s]?phrase\s*[:=]\s*['"\x60]` (i) | Seed phrase assignment |
detectortable -
low Risky intent
intent-wallet-secretsscan-rules.md:118Handles crypto-wallet secrets (seed / mnemonic / private key) — a classic stealer target (detector / deny-list definition; documentation table row)| `recovery[_\s]?phrase\s*[:=]\s*['"\x60]` (i) | Recovery phrase assignment |
detectortable -
low Concealment
en-hide-from-userSKILL.md:761Instruction to hide actions from the user (negated — the text forbids it)- Highlight the most critical risks and explain **why** they matter (e.g. "Your ~/.ssh/ permissions allow any process running as your user to read your private keys, which means a malicious skill coul
negated
A further 12 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
body-longSKILL.md body ≈ 10576 tokens (recommended < 5000); move details to references/ - note
frontmatter-keyunknown frontmatter key "filesystem-access"
Process rating: all ten parameters 44/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 3 mutating operations with no state check
- 40Consistency. Frontmatter name (agentguard) differs from the folder (security)
- 40Execution cost. Instruction body is 10576 tokens: crowds the task out of the window
- 100Tools and files. Tools declared in frontmatter
- 100Steps. 32 steps
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low The response is described with custom markup (59 tags): a typed call is more reliable
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
- -32 of 5 scripts are never mentioned in SKILL.md
- +2Single-language instructions
- +3Description length 385: enough signal without eating the budget
- +4Structure: 28 headings
- +3Step-by-step instructions: 32 items
- +4Has examples (26 code blocks)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 71.