Deep Thinking
v1.0.0Comprehensive deep reasoning framework that guides systematic, thorough thinking for complex tasks. Automatically applies for multi-step problems, ambiguous...
⭐ 1· 102·0 current·0 all-time
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 (deep reasoning for complex tasks) align with the SKILL.md and reference content: guidance on when/how to apply deeper analysis. No unrelated binaries, env vars, or config paths are requested.
Instruction Scope
The instructions are confined to reasoning style and process; they do not direct the agent to read system files, call external endpoints, or access secrets. Note: the protocol explicitly encourages detailed inner-monologue phrasing and exploratory text — that can produce verbose outputs that surface context or user-supplied data in more explicit ways. If you care about avoiding chain-of-thought disclosure or very long responses, consider that this skill encourages that style.
Install Mechanism
No install spec and no code files — instruction-only. Nothing will be downloaded or written to disk by the skill itself.
Credentials
No required environment variables, credentials, or config paths are declared or referenced in the instructions.
Persistence & Privilege
Defaults are used (always:false, model invocation allowed). Autonomous invocation is the platform default and is not combined with other concerning permissions here, so no additional privilege concern.
Assessment
This skill is internally coherent and low-risk: it only provides a reasoning protocol and does not request credentials or install code. Things to consider before enabling: (1) it encourages verbose inner-monologue and exploratory language, which can make replies longer and may expose more contextual detail than a terse answer — avoid if you need concise or redacted outputs; (2) autonomous invocation (the platform default) means the agent may apply this protocol automatically when it judges a task complex — if you want to prevent that, install with model invocation disabled or limit when the skill is used. No evidence of hidden endpoints, file access, or credential requests was found.Like a lobster shell, security has layers — review code before you run it.
latestvk97egn9x4w35mvymhqrvss7d3s83zs46
License
MIT-0
Free to use, modify, and redistribute. No attribution required.