BC bilibili-garb
Bilibili garb (个性装扮) data collection and management. Search garb items, query suit/collection details, scan benefit data for owned items (including discontinued), and determine scarcity tiers. Use when working with Bilibili's personalization system (装扮), digital card collections (收藏集), suit items (套装子项), or any task involving bilibili.com/garb or bilibili.com/h5/mall/digital-card endpoints. Triggers on "bilibili garb", "B站装扮", "装扮搜索", "套装查询", "收藏集", "benefit扫描", "品级判别".
As a process C 57/100 · Has gaps — weak spots: result and completion, inputs and preconditions, failures and branches
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
- 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
net-credential-usescripts/bilibili-garb-collection.sh:103Credential used in a network call (verify the destination is the intended service)SEARCH_IMG=$(curl -s "${API_URL}?key_word=${ENCODED}&pn=1&ps=5&mobi_app=iphone&platform=ios&appkey=${APPKEY}" 2>/dev/null | jq -r '.data.list[0].properties.image_cover // empty' 2>/dev/null) -
medium Exfiltration
net-credential-usescripts/bilibili-garb-collection.sh:180Credential used in a network call (verify the destination is the intended service)RESPONSE=$(curl -s "${API_URL}?key_word=${ENCODED}&pn=1&ps=50&mobi_app=iphone&platform=ios&appkey=${APPKEY}") -
medium Exfiltration
net-credential-usescripts/bilibili-garb-search.sh:39Credential used in a network call (verify the destination is the intended service)RESPONSE=$(curl -s "${API_URL}?key_word=${ENCODED}&pn=1&ps=20&mobi_app=iphone&platform=ios&appkey=${APPKEY}") -
low Exfiltration
exfil-secret-in-urlscripts/bilibili-garb-collection.sh:214Credential passed in a URL query string (normal for some APIs — verify the host is the intended service) (placeholder value)LOTTERY_RESPONSE=$(curl -s "https://api.bilibili.com/x/vas/dlc_act/lottery_home_detail?act_id=${ID}&lottery_id=…&mobi_app=…&platform=ios&appkey=…&access_key=…" 2>/placeholder -
low Exfiltration
net-credential-usescripts/bilibili-garb-collection.sh:214Credential used in a network call (verify the destination is the intended service) (the skill's own vendor host)LOTTERY_RESPONSE=$(curl -s "https://api.bilibili.com/x/vas/dlc_act/lottery_home_detail?act_id=${ID}&lottery_id=…&mobi_app=…&platform=ios&appkey=…&access_key=…" 2>/vendor-host -
low Exfiltration
exfil-secret-in-urlscripts/bilibili-garb-collection.sh:239Credential passed in a URL query string (normal for some APIs — verify the host is the intended service) (placeholder value)BASIC_RESPONSE=$(curl -s "https://api.bilibili.com/x/vas/dlc_act/act/basic?act_id=${ID}&mobi_app=…&platform=ios&appkey=…" 2>/dev/null)placeholder -
low Exfiltration
net-credential-usescripts/bilibili-garb-collection.sh:239Credential used in a network call (verify the destination is the intended service) (the skill's own vendor host)BASIC_RESPONSE=$(curl -s "https://api.bilibili.com/x/vas/dlc_act/act/basic?act_id=${ID}&mobi_app=…&platform=ios&appkey=…" 2>/dev/null)vendor-host -
low Exfiltration
exfil-secret-in-urlscripts/bilibili-garb-collection.sh:265Credential passed in a URL query string (normal for some APIs — verify the host is the intended service) (placeholder value)LOTTERY_RESPONSE=$(curl -s "https://api.bilibili.com/x/vas/dlc_act/lottery_home_detail?act_id=${ID}&lottery_id=…&mobi_app=…&platform=ios&appkey=…" 2>/dev/null)placeholder -
low Exfiltration
net-credential-usescripts/bilibili-garb-collection.sh:265Credential used in a network call (verify the destination is the intended service) (the skill's own vendor host)LOTTERY_RESPONSE=$(curl -s "https://api.bilibili.com/x/vas/dlc_act/lottery_home_detail?act_id=${ID}&lottery_id=…&mobi_app=…&platform=ios&appkey=…" 2>/dev/null)vendor-host
Files scanned: 8. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
✓ No remarks against the Agent Skills spec
Process rating: all ten parameters 57/100
- 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
- 0Progress reporting. Says nothing while it works
- 30Running it twice. 2 mutating operations with no state check
- 70When it triggers. States when to use, but not when not to
- 100Tools and files. No external tools needed
- 100Steps. 15 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 735 tokens
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
- +4Description does not say when NOT to use the skill (false activations)
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +5Description quotes 2 example trigger phrases
- +3Description length 474: enough signal without eating the budget
- +4Structure: 8 headings
- +3Step-by-step instructions: 15 items
- +4Has examples (5 code blocks)
- +4Reference files are cited in the instructions (2 of 2)
- +3All 3 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 96.