harald/zeroclaw
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 3

Fix skill instruction/tool injection in system prompts

Browse source
This commit is contained in:
Chummy 2026-02-20 01:19:43 +08:00
parent f2ffd653de
commit 3733856093
5 changed files with 181 additions and 172 deletions
Download patch file Download diff file Expand all files Collapse all files

2
docs/commands-reference.md
View file

@ -102,6 +102,8 @@ Runtime in-chat commands (Telegram/Discord while channel server is running):
- `zeroclaw skills install <source>`
- `zeroclaw skills remove <name>`
Skill manifests (`SKILL.toml`) support `prompts` and `[[tools]]`; both are injected into the agent system prompt at runtime, so the model can follow skill instructions without manually reading skill files.
### `migrate`
- `zeroclaw migrate openclaw [--source <path>] [--dry-run]`