scienceworld-task-parser
Analyzes user instructions in ScienceWorld environments to extract specific task requirements and constraints. Use when receiving a new task to iden…
它会碰到什么
这一栏是扫描器报的事实,不是结论。命中多不等于有毒(安全工具、规则库、示例脚本本来就会包含危险写法),命中少也不等于干净。它和你手上的凭据、文件、网络有什么关系,需要你自己看。
技能内容
Instructions for ScienceWorld Task Parsing
When you receive a new task instruction in a ScienceWorld environment, follow this process to parse it into a structured plan.
1. Parse the Instruction
Immediately analyze the user's natural language instruction. Extract the following core components:
- Target Object Type: Identify the category or description of the object to be manipulated (e.g., 'non-living thing', 'liquid', 'electrical component').
- Target Location: Identify the final destination for the object, including the room and specific container (e.g., 'purple box in the workshop').
- Required Actions: Infer the sequence of actions implied by verbs like "find", "focus on", "move", "pour", "mix", etc.
Output your analysis as a concise thought. Example: "Task requires finding a non-living object in the workshop and moving it to the purple box."
2. Survey the Environment
- Use
look aroundin your current room to get an inventory of visible objects, containers, and their states. - If the target location is a different room, use
teleport to LOCto go there first, thenlook around.
3. Identify the Target Object
- From the room description, identify objects matching the parsed Target Object Type.
- If multiple candidates exist, select one that is clearly non-living, portable, and not part of a fixed apparatus (e.g., a wire, a light bulb, a battery). Avoid substances like 'air'.
- Use
examine OBJorlook at OBJif you need more detail to confirm an object's properties.
4. Execute the Task Sequence
- Signal Intent: Use
focus on OBJon the identified target object. This explicitly marks the object for the task. - Perform Core Action: Execute the primary action from the parsed instruction (e.g.,
move OBJ to OBJ,pour OBJ into OBJ). - Use
waitorwait1only if necessary to allow for state changes.
Key Principles
- Efficiency: All containers are pre-opened. Do not use
openorcloseunless explicitly required. - Directness: Teleport directly to the target room. Do not explore unrelated rooms.
- Clarity: Structure your internal reasoning using the "Thought:" prefix before each action, as shown in the trajectory.
- Verification: If an initial
look aroundis insufficient, a secondlook aroundis acceptable to confirm the environment state before proceeding.
Example
Task instruction: "Find a non-living thing in the workshop and move it to the purple box."
- Parse: Target = non-living object, Location = workshop, Container = purple box.
teleport to workshoplook around— observe: "a battery, a blue light bulb, an orange wire..."- Select: battery (non-living, portable).
focus on batterymove battery to purple box
想直接用这个技能?
本站把开放许可(MIT / Apache 等)的技能按仓库打包整理到网盘,点一下转存到你自己的网盘,不用一个个从 GitHub 拉。许可未声明的技能只给原始仓库链接,不打包。
它属于哪个仓库
experiments/src/skills/scienceworld/scienceworld-task-parser/SKILL.md同一个仓库里的其他技能
- alfworld-appliance-navigator
- alfworld-appliance-preparer
- alfworld-clean-object
- alfworld-device-operator
- alfworld-environment-scanner
- alfworld-goal-interpreter
- alfworld-heat-object-with-appliance
- alfworld-inventory-management
- alfworld-locate-target-object
- alfworld-location-navigator
- alfworld-navigation-planner
- alfworld-object-cooler