ClawHub

SSQ Prediction

v1.1.0

AI Investment Research - Double Color Ball (双色球) Prediction. Fetches historical data, applies Contextual Quantum-like Bayesian Network (CQ-BN) model, and sen...

⭐ 0· 80·0 current·0 all-time
byRix Zhang@rix-zhang
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Pending
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description (SSQ prediction) align with what the skill asks for: fetching public SSQ history, running a model (described in prose), and optionally sending results to Feishu via the openclaw CLI. The only external requirement is the openclaw binary for message delivery, which matches the delivery capability.
✓
Instruction Scope
SKILL.md instructs the agent to fetch historical data from a public endpoint (datachart.500.com), run an internal CQ-BN simulation described in prompt form, format a report, and optionally call openclaw to send a Feishu message. The instructions do not ask the agent to read unrelated local files, harvest environment secrets, or post data to unexpected external endpoints. The model/computation is described as an LLM role-play (quantum physicist) rather than invoking external code; this is consistent with an instruction-only skill.
✓
Install Mechanism
No install spec or downloadable code is present (instruction-only). This minimizes installation risk — nothing is written to disk by the skill itself.
✓
Credentials
No credentials or sensitive environment variables are required. The SKILL.md documents a single optional variable (FEISHU_SSQ_GROUP_ID) used only to select the Feishu target; that is proportional to the delivery feature. No unrelated secrets or config paths are requested.
✓
Persistence & Privilege
always is false and the skill does not request persistent system-wide privileges or modify other skills. It relies on platform-level Feishu auth handled by OpenClaw and will only send messages when run (or if the optional FEISHU_SSQ_GROUP_ID is set).
Assessment
This skill is coherent: it fetches public historical lottery data, asks the agent (LLM) to simulate a model, and optionally posts the formatted report to a Feishu group via the openclaw CLI. Before installing, confirm: (1) the OpenClaw platform already has Feishu configured and you are comfortable with the skill posting to the specified group (only set FEISHU_SSQ_GROUP_ID if you want automatic posting), (2) your environment allows outbound HTTP requests to the datachart.500.com endpoint, and (3) you understand the prediction is speculative — the SKILL.md is a prompt-driven model description rather than audited code, so there is no guarantee of scientific validity. Because the skill is instruction-only and asks for no secrets, the security risk is low; if you need stronger assurance, request the author provide an auditable implementation or explicit details of the numerical algorithm used.

Like a lobster shell, security has layers — review code before you run it.

latestvk974z17yjnjhwwjqb3tdh5wmfd83qvxq

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Runtime requirements

Binsopenclaw
Environment variables
FEISHU_SSQ_GROUP_IDoptional— Feishu group ID to deliver the prediction report. If not set, output is shown in chat.

Comments