Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Hub Api Integration
v1.0.0Help apps and distribution channels integrate PCS Hub into their frontend. Use when user says "/hub-api-integration", "integrate PCS Hub", "embed PCS Hub swa...
⭐ 0· 46·0 current·0 all-time
by@pcs-bot
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
high confidencePurpose & Capability
The stated purpose—producing an integration spec for embedding PCS Hub—matches the instruction-only design. Required binaries (curl, jq) and the optional use of cast/foundry for blockchain queries are plausible. However, installing Foundry/ cast is a heavier dependency than a pure spec generator needs; it may be justified if on-chain queries (route/quote verification) are expected but this isn't made explicit in the registry metadata.
Instruction Scope
SKILL.md contains detailed runtime instructions: a telemetry-style curl 'ping' to pancake.ai that sends agent identifier, OS/arch, and timestamp; many steps to collect user requirements; and explicit security rules that reference reading a Hub API token from the environment. The instructions do not ask to read arbitrary system files, but the presence of a telemetry call and requirements to access an (undeclared) auth token expand the runtime scope beyond 'just produce a spec'.
Install Mechanism
Install uses brew formulas for curl, jq, and foundry. These are standard package sources on macOS/Linux and not inherently suspicious. The inclusion of foundry (cast) is notable because it pulls developer tooling; that's reasonable if on-chain queries are needed but is heavier than strictly necessary for generating documentation/specs.
Credentials
The SKILL.md explicitly references a sensitive Hub API token (PCS_HUB_TOKEN), instructing agents to read it from the environment and never print it, yet the registry metadata lists no required environment variables or primary credential. This is a clear mismatch. The skill also refers to using public BSC RPC endpoints (which may require RPC URLs/keys) but does not declare any ENV requirements for RPC access. Asking for an undisclosed secret is disproportionate and should be fixed or explained.
Persistence & Privilege
The skill is not set to always:true and does not request system-wide config paths or persistent installation beyond optional brew installs. Allowed tools include file read/write/edit but that's consistent with producing spec deliverables. No elevated persistent privileges are requested.
What to consider before installing
This skill largely does what it says (produce a PCS Hub integration spec), but there are two red flags you should get clarified before installing: (1) SKILL.md expects a sensitive environment variable named PCS_HUB_TOKEN but the skill metadata does not declare it — ask the author why it's needed and where it will be stored/read; (2) the skill performs a background 'ping' to https://pancakeswap.ai with your agent name, OS, arch and timestamp — decide whether you are comfortable with that telemetry. Also confirm whether heavy developer tooling (Foundry/cast) is actually required for your use case. If the author cannot justify the undeclared secret or telemetry, avoid installing or require that the skill be updated to explicitly declare required env vars and to make telemetry optional/explicit.Like a lobster shell, security has layers — review code before you run it.
latestvk97fj7zt5335bfdw9vpv7yr9rx83hxts
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
OSmacOS · Linux
Binscurl, jq
Any bincast, python3, node, open, xdg-open
Install
Homebrew
Bins: curl
brew install curlHomebrew
Bins: jq
brew install jqHomebrew
Bins: cast
brew install foundry