Skip to content
agent-slides
Docs Examples Notes GitHub
Notes

The thinking behind agent-slides.

Design decisions, architecture patterns, and lessons learned building an agent skill for PowerPoint generation.

01

Why agents need a different CLI

Human DX and Agent DX are different problems. Why the agent needs a dedicated execution layer, not raw library access.

Read →
02

JSON payloads and runtime schemas

Why CLIs for agents should accept structured data instead of flags, and why the API contract should come from the tool itself.

Read →
03

Context-window discipline

Why output volume is a hard design constraint, and why deterministic output is non-negotiable for automation.

Read →
04

Agent output as untrusted input

Why agent-driven CLIs should validate like a public API, and why design taste must become machine-readable contracts.

Read →
05

Skills encode expertise, not instructions

Why --help isn't enough. Skill files give agents the workflow knowledge that documentation can't convey.

Read →
06

Why seven skills, not one

Why one monolithic skill became seven focused ones, and why conditional reference loading keeps context lean.

Read →
07

File contracts and state-machine orchestration

Why skills communicate through files on disk, and why the orchestrator converges through retry loops instead of a linear pipeline.

Read →
08

Progressive disclosure for agents

Give the agent a map, not a manual. Each layer reveals only what the current step needs, so context stays lean and decisions stay grounded.

Read →
agent-slides

An agent skill for PowerPoint generation.
Open-source. Python 3.12+.

Learn
Docs Examples Notes
Links
GitHub