WHEN HOTLINE
Choose Hotline when...
When the capability's core implementation needs protection, monetization, or has high execution cost that shouldn't be borne by model context.
Use Hotline
KEY TAKEAWAYS
In three lines
- Skills guide behavior inside model context; Hotline provides callable capabilities outside context — different layers.
- Skills can call Hotlines as an execution backend — they compose naturally.
- Commercialization, versioning, ratings, and settlement need to be built outside Skills; Hotline provides that layer.
DIMENSIONS
Dimension by dimension
Each row compares what Hotline standardizes, what Agent Skills focuses on, and why the difference matters.
Dimension
Hotline
Agent Skills
Execution location
Runs at the Responder side; model sees only input/output schema.
Why it matters
Execution location determines whether implementation logic needs to enter the prompt.
Great for injecting instructions, workflows, and preferences into model context for in-context execution.
Prompt footprint
Near-zero prompt footprint beyond discovery metadata.
Why it matters
Affects how many capabilities an agent can carry in long conversations.
Consumes prompt space and token budget; well-suited for lightweight behavioral guidance.
Commercial model
Per-call settlement at the protocol layer.
Why it matters
Capabilities that need monetization require protocol-level settlement.
Great for sharing behavioral guidance freely; commercialization, versioning, ratings, and settlement need to be built externally.
Source visibility
Implementation remains black-boxed behind a contract.
Why it matters
Different visibility models serve different purposes.
Instruction logic is visible to the model — by design, for behavior steering.
WHEN AGENT SKILLS
Choose Agent Skills when...
When your goal is to steer model behavior — injecting instructions, workflow preferences, and process guidance into context. Skills are highly efficient at behavior injection and workflow guidance.
Not necessarily Hotline
PAGE FAQ
Common questions about Agent Skills
Can a Skill call a Hotline?
Yes, and it's a recommended composition pattern. The Skill teaches the model when and why to call a capability; the Hotline handles the actual execution and settlement.
Does Hotline replace every use of Skills?
No. Skills are still the right tool for instruction, orchestration hints, and local behavior shaping. Hotline becomes valuable when the capability itself is the product or needs runtime isolation.
What's the best way to combine them?
Use a Skill to tell the model a capability exists and when to invoke it; use the Hotline for the protected execution and settlement. A Hotline can ship an official Skill template that callers drop into their project.