You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Default lint mode focuses on deterministic publish-readiness problems. Use `--strict` to include advisory checks such as trigger-description quality, large skill bodies, unreferenced scripts, and plugin name/folder mismatch.
188
188
189
+
## Configuration
190
+
191
+
Add `skillforge.json` at the repo or package root to tune checks:
192
+
193
+
```json
194
+
{
195
+
"name": "my-agent-skill-pack",
196
+
"lint": {
197
+
"ignore": [
198
+
"templates/**",
199
+
"tests/fixtures/**",
200
+
"examples/broken-on-purpose/**"
201
+
]
202
+
},
203
+
"checks": {
204
+
"maxSkillMdLines": 500,
205
+
"requireOpenAiYaml": false,
206
+
"allowScripts": true
207
+
}
208
+
}
209
+
```
210
+
211
+
Use `lint.ignore` for intentional fixtures, vendored examples, generated output, or template folders that should not be treated as publishable skills/plugins.
212
+
189
213
## Compatibility
190
214
191
215
Use `compat` to check whether a skill/package is likely to work in a specific agent ecosystem.
0 commit comments