Unverified50% confidenceFactExact time
The Agent retains only a single Bash tool, and all capabilities must be implemented through shell commands.
1
Sources
50%
Confidence
Long-term
Relevance
9/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAgent Skill adopts a lazy-loading mechanism whereby the Agent first reads only the metadata (name and description) of each Skill and loads the full instructions only after matching the user's need, so having multiple Skills coexisting does not increase Token consumption70% similarUnverifiedIn Agent systems, permissions should be independently verified for each tool call rather than relying on session-level authorization, combined with the principle of least privilege to revoke permissions immediately after an operation is complete.68% similarUnverifiedThe General Purpose Agent is the only core Agent with full read, write, and run-command permissions, and it loads CLAUDE.MD.66% similarUnverifiedThe "Capability Token" mechanism is being explored by some frameworks, requiring Agents to hold a one-time authorization token each time they invoke a high-risk tool.65% similarUnverified代理在启动时只加载每个技能的名称和描述，而非全部内容，以避免上下文窗口爆掉65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/865022API
curl https://kongchang.com/api/v1/knowledge/claims/865022MCP
get_claim(id=865022)