Skip to content

Commit 4a3f0de

Browse files
ci: apply automated fixes
1 parent 78841cd commit 4a3f0de

File tree

1 file changed

+6
-6
lines changed

1 file changed

+6
-6
lines changed

src/blog/from-docs-to-agents.md

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -1,14 +1,14 @@
11
---
2-
title: "From Docs to Agents: Shipping Knowledge That Travels With Your Package"
2+
title: 'From Docs to Agents: Shipping Knowledge That Travels With Your Package'
33
published: 2026-03-04
44
authors:
55
- Sarah Gerrard
6-
- Kyle Matthews
6+
- Kyle Matthews
77
---
88

99
Your docs are good. Your types are solid. Your agent still gets it wrong.
1010

11-
Not because it's dumb — because no pipeline connects what you know about your tool to what agents know. Docs target humans who browse. Types check individual API calls but can't encode intent. Training data snapshots the ecosystem as it *was*, mixing versions without flagging which one applies. The gap isn't content. It's lifecycle.
11+
Not because it's dumb — because no pipeline connects what you know about your tool to what agents know. Docs target humans who browse. Types check individual API calls but can't encode intent. Training data snapshots the ecosystem as it _was_, mixing versions without flagging which one applies. The gap isn't content. It's lifecycle.
1212

1313
## Skills as side quests
1414

@@ -72,7 +72,7 @@ npx intent validate
7272
npx intent setup
7373
```
7474

75-
The alternative — hoping model providers re-train on your latest docs — is not a strategy. Training data has a permanent version-mixing problem: once a breaking change ships, models contain *both* versions forever with no way to disambiguate. Skills bypass this entirely. They're versioned with your package, and `npm update` brings the latest knowledge with the latest code.
75+
The alternative — hoping model providers re-train on your latest docs — is not a strategy. Training data has a permanent version-mixing problem: once a breaking change ships, models contain _both_ versions forever with no way to disambiguate. Skills bypass this entirely. They're versioned with your package, and `npm update` brings the latest knowledge with the latest code.
7676

7777
![Model training data mixes versions permanently vs. skills pinned to your installed version](./diagram-split-brain.svg)
7878

@@ -103,7 +103,7 @@ npx intent meta
103103

104104
## From skills to playbooks
105105

106-
A single skill helps an agent use one tool correctly. But real development demands composition — routing *with* server state *with* a data grid *with* client-side storage. No individual skill covers how they fit together.
106+
A single skill helps an agent use one tool correctly. But real development demands composition — routing _with_ server state _with_ a data grid _with_ client-side storage. No individual skill covers how they fit together.
107107

108108
Playbooks are the orchestration layer. A developer says "build a paginated data table with URL-synced filters" and the playbook loads the right skills in the right order — search params, loader/query integration, table columnDefs. Developer goals map to skill combinations.
109109

@@ -138,4 +138,4 @@ Devtool makers have a new surface to maintain. You used to ship code, docs, and
138138

139139
Tools that invest here will produce developers who build confidently from day one — not through tutorials or toy projects, but through correct patterns absorbed in the context of real work.
140140

141-
The lifecycle is: write your docs, generate skills, ship them with your package, validate and keep them current, learn from how they're used, make your tool better. Repeat.
141+
The lifecycle is: write your docs, generate skills, ship them with your package, validate and keep them current, learn from how they're used, make your tool better. Repeat.

0 commit comments

Comments
 (0)