← Back to Skills Marketplace
1093
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install prediction-market
Description
Participate in Base's GTM by betting if today's 18:00 UTC garden temperature will be higher or lower than yesterday's, sharing winnings from the pot.
Usage Guidance
This skill appears to be a straightforward helper for a Base-based temperature prediction market. Before installing or using it: (1) do not paste your private key into untrusted places — the instructions require a signing key to send bets (cast/Bankr), so use a secure wallet or signing service; (2) verify the contract addresses and keeper/trust model yourself — the README warns the keeper submits settlement values off-chain; (3) the bundle contains Foundry/forge-std helper files (including a script that fetches cheatcodes from GitHub if you run it) — only run included scripts if you trust their behavior and source; (4) avoid reusing sensitive credentials (PRIVATE_KEY, BASESCAN_API_KEY) unless you control them and understand the transaction consequences. If you want additional assurance, ask the skill author for a canonical homepage/repo or an audit of the contract and keeper workflow.
Capability Analysis
Type: OpenClaw Skill
Name: prediction-market
Version: 1.0.1
The skill bundle is classified as suspicious due to the presence of shell commands in `SKILL.md` that directly handle sensitive data (private keys) via the `--private-key $KEY` argument for `cast send`. While this is a common pattern for CLI tools and is plausibly needed for blockchain interaction, it represents a significant vulnerability surface if the AI agent's execution environment or prompt handling is not robust against shell injection or insecure credential management. Additionally, the `lib/forge-std/foundry.toml` file contains broad filesystem write permissions (`fs_permissions = [{ access = "read-write", path = "./" }]`) which, although intended for development/testing, could be exploited if the agent were to execute arbitrary `forge` commands. There is no clear evidence of intentional malicious behavior, but these risky capabilities warrant a 'suspicious' classification.
Capability Assessment
Purpose & Capability
Name, files, and SKILL.md all describe interacting with a daily temperature prediction market on Base (contract address, RPC, cast examples, Bankr payloads). The included source, README, and PLAYERS.md are consistent with that purpose. No unrelated credentials or binaries are requested.
Instruction Scope
Runtime instructions are narrowly focused on reading contract state (cast call) and submitting bets (cast send or Bankr JSON). They reference using a private key (e.g. $KEY / YOUR_PRIVATE_KEY) to sign transactions — expected for on‑chain actions but not declared in requires.env. The README explicitly warns about the trusted keeper model. No instructions ask the agent to read arbitrary host files or exfiltrate data.
Install Mechanism
No install spec is provided (instruction-only skill at runtime), so nothing is automatically downloaded or executed by an installer. The repo includes Foundry/forge-std files and scripts, but those are static file contents in the bundle — they are not executed unless a user/agent explicitly runs them.
Credentials
The skill declares no required environment variables (none in requires.env), which is proportionate. However SKILL.md and README show usage examples that require a private key (PRIVATE_KEY, $KEY) and a Basescan API key for deployment — the skill does not declare these, so a user/agent must supply them externally to place bets or deploy. This is expected for blockchain interaction but is an important user responsibility.
Persistence & Privilege
The skill does not request always:true, does not declare persistence or modify other skills, and has normal autonomous-invocation settings. There is no evidence it attempts to persist credentials or modify global agent config.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install prediction-market - After installation, invoke the skill by name or use
/prediction-market - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
Rebranded to Garden Temp Market (GTM)
v1.0.0
- Initial release: DailyTempMarket lets users bet on whether the daily 18:00 UTC temperature on Base will be higher or lower than the previous day.
- Supports full market state queries, individual data endpoints, and bet placement via Bankr API or cast.
- Details game rules, schedule (betting closes at 12:00 UTC, settles 18:00 UTC), and agent strategy example.
- Supports up to 200 bettors per side, 0.001 ETH minimum bet, and automatic rollovers for ties.
- Integrates temperature data from SensorNet; full event interface provided for tracking bets, settlements, and claims.
Metadata
Frequently Asked Questions
What is Garden Temp Market (GTM)?
Participate in Base's GTM by betting if today's 18:00 UTC garden temperature will be higher or lower than yesterday's, sharing winnings from the pot. It is an AI Agent Skill for Claude Code / OpenClaw, with 1093 downloads so far.
How do I install Garden Temp Market (GTM)?
Run "/install prediction-market" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Garden Temp Market (GTM) free?
Yes, Garden Temp Market (GTM) is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Garden Temp Market (GTM) support?
Garden Temp Market (GTM) is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Garden Temp Market (GTM)?
It is built and maintained by Potdealer (@potdealer); the current version is v1.0.1.
More Skills