Strategy Builder
v1.0.0Design, backtest, and reason about trading strategies for Paradex using MCP tools. Takes natural language strategy descriptions and turns them into structure...
⭐ 0· 33·0 current·0 all-time
bySergey Vidyuk@sv
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (Paradex strategy design and backtesting) matches the instructions: the SKILL.md describes building strategy specs, using Paradex MCP data tools (klines, trades, orderbook, funding, markets) for validation and execution planning. There are no unrelated binaries, credentials, or installs requested.
Instruction Scope
SKILL.md confines the agent to design, signal generation, and sanity-check style historical validation. It explicitly states it will not execute trades and only references Paradex MCP data tools and market metadata. It does not instruct the agent to read arbitrary system files, environment variables, or to transmit data to external endpoints outside the platform tools.
Install Mechanism
No install specification and no code files — instruction-only. That reduces disk and network risk; nothing is downloaded or executed beyond platform tool calls.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. It references Paradex tools that are expected for market-data access; asking for no secrets is proportionate to a design/backtest skill.
Persistence & Privilege
The skill does not request always:true and has default autonomous invocation allowed. Autonomous invocation is normal for skills; if you are concerned about an agent acting without prompts, consider disabling autonomous invocation at the agent/policy level — the skill itself does not request elevated persistence or system-wide configuration changes.
Assessment
This skill appears coherent and limited to strategy design and light backtesting using Paradex MCP data. Before installing: (1) Confirm you understand that it will not execute trades — execution requires separate authenticated order-management tools and credentials. (2) Verify your platform's MCP tool permissions and auditing (the skill will call paradex_klines, paradex_trades, etc.). (3) Do not provide credentials to this skill (it doesn't ask for any); if you later connect an execution-capable skill, treat those credentials carefully. (4) If you want to reduce risk, disable autonomous invocation for the agent or require explicit confirmation before any MCP calls. (5) Remember the backtests described are sanity checks (the SKILL.md notes they don't account for fills/slippage/real execution).Like a lobster shell, security has layers — review code before you run it.
latestvk972p822kesepnt1tfp3277amx840wk8
License
MIT-0
Free to use, modify, and redistribute. No attribution required.