BB home-assistant-toolkit
Home Assistant full management skill — control devices, create automations, monitor health, manage backups, update HACS, generate dashboards, all via SSH. Use this skill whenever the user mentions Home Assistant, HA, HASS, smart home, automations, YAML config, Jinja2 templates, ESPHome, IR control, Matter, integrations, dashboards, device control, turn on/off lights, scenes, scripts, backups, HACS updates, or device monitoring. Connects to HA via SSH for full access: config files, ha CLI, REST API, logs. Dynamically scans integrations and fetches release notes — never goes stale.
As a process B 70/100 · Nearly there — weak spots: result and completion
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.
How to improve
- 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 · 7
✓ No critical or high findings
Medium and low: 7
-
medium Exfiltration
net-credential-usescripts/check-setup.sh:29Credential used in a network call (verify the destination is the intended service)if result=$(curl -sf -m 5 -H "Authorization: Bearer ${HA_TOKEN}" "${HA_URL}/api/config" 2>/dev/null); then -
medium Exfiltration
net-credential-usescripts/common.sh:124Credential used in a network call (verify the destination is the intended service)curl -sf -H "Authorization: Bearer ${HA_TOKEN}" -H "Content-Type: application/json" "${HA_URL}/api/${endpoint}" -
medium Exfiltration
net-credential-usescripts/common.sh:126Credential used in a network call (verify the destination is the intended service)curl -sf -X "$method" -H "Authorization: Bearer ${HA_TOKEN}" -H "Content-Type: application/json" -d "${data}" "${HA_URL}/api/${endpoint}" -
low Exfiltration
net-credential-usescripts/common.sh:138Credential used in a network call (verify the destination is the intended service) (quoted — discussed, not commanded)ssh_cmd "curl -sf -H \"Authorization: Bearer ${token_ref}\" -H 'Content-Type: application/json' \"http://supervisor/core/api/${safe_endpoint}\""quoted -
low Exfiltration
net-credential-usescripts/common.sh:142Credential used in a network call (verify the destination is the intended service) (quoted — discussed, not commanded)ssh_cmd "curl -sf -X ${safe_method} -H \"Authorization: Bearer ${token_ref}\" -H 'Content-Type: application/json' -d '${safe_data}' \"http://supervisor/core/api/${safe_endpoint}\""quoted -
low Exfiltration
net-credential-usescripts/ha.sh:116Credential used in a network call (verify the destination is the intended service) (quoted — discussed, not commanded)raw_addons=$(ssh_cmd "curl -sf -H \"Authorization: Bearer ${sv_token_ref}\" http://supervisor/addons" 2>/dev/null || echo "")quoted -
low Exfiltration
net-credential-usescripts/ha.sh:303Credential used in a network call (verify the destination is the intended service) (quoted — discussed, not commanded)raw=$(ssh_cmd "curl -sf -H \"Authorization: Bearer ${sv_token_ref}\" -H \"Content-Type: application/json\" http://supervisor/addons")quoted
Files scanned: 17. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 5606 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 70/100
- 0Result and completion. Does not say what the result is
- 60Tools and files. Uses tools (bash, web, python) that frontmatter does not declare
- 70When it triggers. States when to use, but not when not to
- 70Inputs and preconditions. Inputs and preconditions are listed
- 70Execution cost. Instruction body is 5606 tokens
- 100Steps. 95 steps
- 100Failures and branches. 12 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 11 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)
- +3Output format is not stated: the model decides each time
- -31 of 11 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +3Description length 586: enough signal without eating the budget
- +4Structure: 37 headings
- +3Step-by-step instructions: 95 items
- +4Has examples (24 code blocks)
- +4Reference files are cited in the instructions (2 of 2)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 75.