Unverified50% confidenceFactExact time
WeAct trains multiple capability-specific LoRA adapters on a frozen backbone model, with each adapter corresponding to one capability; adapter selection is modeled as an internal action of the agent, on par with calling external tools.
1
Sources
50%
Confidence
Long-term
Relevance
10/20/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWeAct proposes adding capability-specific LoRA adapter selection to the agent's action space, achieving on-demand capability composition through the native tool-calling interface.80% similarUnverifiedModeling handoff as a tool call reuses the model's already well-trained tool-selection ability, without requiring the model to learn new collaboration semantics.68% similarVerifiedThe Skill mechanism uses on-demand loading: it first sends the metadata of all Skills to the large model, and after the large model selects the matching Skill, the full instructions are loaded.66% similarUnverifiedThe design of the harness directly determines how many model calls are made to complete a task and how many tokens of context each call carries.66% similarUnverifiedThe tool-calling mechanism decouples the model's intent from actual system operations, allowing developers to insert permission checks, sandbox isolation, or manual confirmation steps in the intermediate layer.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1073253API
curl https://kongchang.com/api/v1/knowledge/claims/1073253MCP
get_claim(id=1073253)