Auto Skill Trigger
v1.0.0Auto Skill Trigger - Match skills to tasks using keyword scoring
⭐ 0· 37·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description match the implementation: the SKILL.md shows a small keyword-overlap algorithm that scores available skills and returns a short prompt hint. It doesn't request unrelated credentials or binaries. It does assume runtime APIs like api.registerHook and getAllSkills exist in the agent environment — reasonable for a hook-style skill but may fail if those APIs are absent.
Instruction Scope
Instructions are limited to scoring skills' descriptions against the incoming message and adding a 'Skill Hint' block to the prompt. They do not read files, env vars, or external endpoints. Caveats: the hint biases the LLM toward listed skills (intentional but may hide relevant skills), the sample code has minor robustness issues (e.g., dividing by keywords.size could be problematic if empty), and it assumes access to the agent's skill registry and SKILL.md content.
Install Mechanism
No install spec and no code files — instruction-only. This is lowest-risk from an installation perspective (nothing is written to disk or downloaded).
Credentials
Requires no environment variables, credentials, or config paths. The only access it needs is the agent-provided ability to list available skills and their descriptions, which is proportional to its purpose.
Persistence & Privilege
always is false and the skill does not request elevated privileges. It registers a 'before_prompt_build' hook (intended behavior) which will influence skill selection when run. Autonomous invocation is allowed by platform default; combined with other concerns that could be significant, but by itself this is a normal pattern for a skill of this type.
Assessment
This skill is coherent and low-risk: it simply ranks available skills by keyword overlap and inserts a short 'Focus on:' hint into the prompt. Before installing, confirm your agent runtime exposes the required hook APIs (api.registerHook, getAllSkills), and be aware that the hint intentionally biases the LLM toward the listed skills (which can hide other relevant skills). The sample code is minimal and has small robustness issues (e.g., no handling for empty keyword sets); consider testing and tuning threshold/maxSkills/stopWords. No secrets or external downloads are required.Like a lobster shell, security has layers — review code before you run it.
latestvk977mc2nhv8m6jrxr0q0cfd8yd842kpz
License
MIT-0
Free to use, modify, and redistribute. No attribution required.