BB kami-suspicious-person
Detect unregistered faces loitering in sensitive areas. Supports one OR many RTSP cameras concurrently in a single process (shared ONNX models, per-camera tracker). Runs continuously, outputs alarm JSON to stdout each time a stranger exceeds the loiter threshold, then keeps monitoring. No local GPU needed for face detection (CPU inference via ONNX).
As a process B 76/100 · Nearly there — weak spots: when it triggers
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 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 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.
How to improve
- 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 · 7
✓ No critical or high findings
Medium and low: 7
-
medium Dangerous commands
cmd-pipe-to-shell-known-hostsetup.sh:44Pipe-to-shell installer from a well-known host (still executes remote code)curl -LsSf https://astral.sh/uv/install.sh | sh
-
medium Dangerous commands
cmd-pipe-to-shell-known-hostsetup.sh:46Pipe-to-shell installer from a well-known host (still executes remote code)wget -qO- https://astral.sh/uv/install.sh | sh
-
medium Exfiltration
exfil-webhook-urlsuspicious_person_detector.py:773Webhook / callback URL commonly used for exfiltration (verify the destination) (quoted — discussed, not commanded)url = f"https://api.telegram.org/bot{bot_token}/sendPhoto"quoted -
medium Exfiltration
exfil-webhook-urlsuspicious_person_detector.py:785Webhook / callback URL commonly used for exfiltration (verify the destination) (quoted — discussed, not commanded)url = f"https://api.telegram.org/bot{bot_token}/sendMessage"quoted -
low Exfiltration
exfil-webhook-urlREADME.md:299Webhook / callback URL commonly used for exfiltration (verify the destination) (placeholder value)4. Click **Copy Webhook URL** (format: `https://discord.com/api/webhooks/123456/abcdef...`)
placeholder -
low Exfiltration
exfil-webhook-urlREADME.md:319Webhook / callback URL commonly used for exfiltration (verify the destination) (placeholder value)- Or call `https://api.telegram.org/bot<TOKEN>/getUpdates` after sending a message in the group → find `"chat":{"id":-100xxxxx}` in the responseplaceholder -
low Exfiltration
exfil-webhook-urlREADME.md:331Webhook / callback URL commonly used for exfiltration (verify the destination) (placeholder value)--discord_webhook https://discord.com/api/webhooks/... \
placeholder
Files scanned: 8. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- note
frontmatter-keyunknown frontmatter key "triggers"
Process rating: all ten parameters 76/100
- 20When it triggers. No condition that starts the skill
- 60Tools and files. Uses tools (bash, web, python) that frontmatter does not declare
- 60Result and completion. Output format stated, no completion criterion
- 70Execution cost. Instruction body is 4312 tokens
- 100Steps. 39 steps
- 100Inputs and preconditions. Inputs and preconditions are listed
- 100Failures and branches. 10 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 14 top-level sections: this looks like several domains in one skill
- low The response is described with custom markup (11 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)
- +1No license
- +2Single-language instructions
- +3Description length 351: enough signal without eating the budget
- +4Structure: 15 headings
- +3Step-by-step instructions: 39 items
- +3Output format is stated explicitly
- +4Has examples (7 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 86.