CD web-access
为AI Agent提供完整的互联网访问能力，支持多个主流平台
As a process D 39/100 · Unfinished process — weak spots: result and completion, when it triggers, inputs and preconditions
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 instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".
If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.
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
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
- 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 · 9
✓ No critical or high findings
Medium and low: 9
-
medium Exfiltration
intent-browser-credential-storecli.py:69Accesses a browser credential / cookie store (quoted — discussed, not commanded)help="Auto-extract ALL platform cookies from browser (chrome/firefox/edge/brave/opera)")
quoted -
medium Exfiltration
intent-browser-credential-storecli.py:200Accesses a browser credential / cookie store (quoted — discussed, not commanded)print("🍪 [dry-run] Would try to import cookies from Chrome/Firefox")quoted -
medium Dangerous commands
cmd-install-from-urlcli.py:820Installs a package from an untrusted URL / archive (string literal in code, not executed)print(" pip install --upgrade https://github.com/Panniantong/agent-reach/archive/main.zip")code literal -
medium Dangerous commands
cmd-install-from-urlcli.py:839Installs a package from an untrusted URL / archive (string literal in code, not executed)print(" pip install --upgrade https://github.com/Panniantong/agent-reach/archive/main.zip")code literal -
medium Dangerous commands
cmd-install-from-urlcli.py:913Installs a package from an untrusted URL / archive (string literal in code, not executed)print(f" 更新: pip install --upgrade https://github.com/Panniantong/agent-reach/archive/main.zip")
code literal -
medium Dangerous commands
cmd-shell-rcinstall.sh:119Writes to a shell startup fileecho "$export_file" >> ~/.bashrc
-
low Dangerous commands
cmd-pipe-to-shell-known-hostcli.py:331Pipe-to-shell installer from a well-known host (still executes remote code) (string literal in code, not executed)"curl -fsSL https://deb.nodesource.com/setup_22.x | bash - 2>/dev/null && apt-get install -y -qq nodejs 2>/dev/null",
code literal -
low Dangerous commands
cmd-pipe-to-shellcli.py:412Downloads and executes remote code from an unrecognised host (pipe to shell) (detector / deny-list definition; string literal in code, not executed)("Node.js", ["node"], "curl NodeSource setup | bash + apt install nodejs"),detectorcode literal -
low Dangerous commands
cmd-pipe-to-shellcli.py:437Downloads and executes remote code from an unrecognised host (pipe to shell) (detector / deny-list definition; string literal in code, not executed)print(" https://nodejs.org/ or: curl -fsSL https://fnm.vercel.app/install | bash")detectorcode literal
Files scanned: 38. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
description-shortdescription under 40 chars: too little signal for triggering - warning
description-no-whendescription does not say WHEN to use the skill (no "use when") - note
frontmatter-keyunknown frontmatter key "requires"
Process rating: all ten parameters 39/100
- 0Result and completion. Does not say what the result is
- 0When it triggers. No condition that starts the skill
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 40Consistency. Frontmatter name (web-access) differs from the folder (07)
- 60Tools and files. Uses tools (python) that frontmatter does not declare
- 100Steps. 33 steps
- 100Execution cost. Instruction body is 961 tokens
- 100Running it twice. No mutating operations
- low 10 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 30: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +4Structure: 31 headings
- +3Step-by-step instructions: 33 items
- +4Has examples (17 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 60.