fix(skills): distinguish native vs external tool credential storage
This commit is contained in:
parent
1faf9d2020
commit
1e0181c7cd
@ -100,14 +100,17 @@ Files not intended to be loaded into context, but rather used within the output
|
||||
|
||||
#### Secrets & Credentials
|
||||
|
||||
**NEVER hardcode secrets automatically**—only if user explicitly requests it. Scripts must look up secrets dynamically: config → env → error.
|
||||
**NEVER hardcode secrets automatically.** Look up secrets dynamically based on skill type:
|
||||
|
||||
**Clawdbot-native skills** (no external CLI): Use config → env → error:
|
||||
```bash
|
||||
VALUE=$(jq -r '.skills.entries["skill-name"].apiKey // empty' ~/.clawdbot/clawdbot.json)
|
||||
VALUE="${VALUE:-$SKILL_NAME_API_KEY}"
|
||||
[[ -z "$VALUE" ]] && echo "Error: Set skills.entries.skill-name.apiKey in config or SKILL_NAME_API_KEY env var" && exit 1
|
||||
```
|
||||
|
||||
**Skills wrapping external tools**: Source from `~/.config/<tool>/` (XDG convention). If the tool works standalone without Clawdbot, its credentials belong outside Clawdbot's config.
|
||||
|
||||
#### What to Not Include in a Skill
|
||||
|
||||
A skill should only contain essential files that directly support its functionality. Do NOT create extraneous documentation or auxiliary files, including:
|
||||
|
||||
Loading…
Reference in New Issue
Block a user