SKILL reference not update after edit the SKILL #19050

Activity

1. Afauria

   added

   bugSomething isn't workingSomething isn't working

   on Mar 25, 2026
2. github-actions

   assigned

   rekram1-node

   on Mar 25, 2026
3. github-actions

   added

   coreAnything pertaining to core functionality of the application (opencode server stuff)Anything pertaining to core functionality of the application (opencode server stuff)

   on Mar 25, 2026
4. ambient-code

   added a commit that references this issue on Mar 25, 2026

   fix(skill): invalidate skill cache when SKILL.md files change on disk

   ...

   This issue will close once commit b35b599 is merged into the 'dev' branch.

   b35b599
5. jwm4 commented on Mar 25, 2026

   

   jwm4

   on Mar 25, 2026

   More actions

   I took a stab at this with draft PR #19136, but after manual testing I found the fix doesn't actually solve the problem for the default configuration.

   What the PR does: Adds a Skill.invalidate() method that clears the cached skills, and wires it to FileWatcher.Event.Updated so that editing a SKILL.md triggers a cache reload. The Command cache also rebuilds when skills are invalidated.

   Why it doesn't work in practice: FileWatcher only watches Instance.directory when OPENCODE_EXPERIMENTAL_FILEWATCHER=true (see src/file/watcher.ts:123). Without that flag, only .git/HEAD is watched, so FileWatcher.Event.Updated never fires for SKILL.md changes. Most users won't have the experimental flag enabled.

   There's also a secondary gap: even with the flag on, the watcher only covers CWD. Skills loaded from parent directories (monorepo .claude/skills/) or global directories (~/.claude/skills/) wouldn't trigger invalidation.

   The cache invalidation logic itself works — Skill.invalidate() correctly clears the cache and the next Skill.all()/Skill.get() call re-reads from disk. All automated tests pass. The gap is in how invalidation gets triggered.

   Possible approaches a maintainer might consider:

   1. Dedicated watcher for skill directories — After the initial skill load, set up fs.watch or parcel watcher subscriptions on just the known skill directories (state.dirs). Independent of the experimental flag.
   2. Poll-based mtime checks — On each ensure() call, stat the loaded SKILL.md files and invalidate if any mtime has changed. Simple but adds disk I/O per call.
   3. Broaden the experimental watcher scope — Extend FileWatcher to also cover skill directories when enabled, and eventually promote it out of experimental.

   Happy to update the PR if a maintainer has a preferred direction.

   👍React with 👍1Reacted by Akiyama
6. abread commented on Apr 10, 2026

   

   abread

   on Apr 10, 2026

   More actions

   Is the skill cache even useful? I doubt the increased memory usage and complexity is worth it.

   What extra information does opencode have about skill usage patterns that can make it do a better job than whatever LRU approximation the OS is already doing for us?

   👍React with 👍3Reacted by CZ, Matt Lyons and Łukasz Wasylkowski
7. tomzx

   marked Skills aren't always refreshed when SKILL.md files are updated #23645 as a duplicate of this issue on Apr 21, 2026
8. thdxr

   removed

   bugSomething isn't workingSomething isn't working

   coreAnything pertaining to core functionality of the application (opencode server stuff)Anything pertaining to core functionality of the application (opencode server stuff)

   on May 3, 2026
9. github-actions commented on Jul 3, 2026

   

   github-actionsBot

   on Jul 3, 2026 – with GitHub Actions

   Contributor

   More actions

   To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.
10. github-actions

    closed this as not plannedon Jul 3, 2026