agent toolingby openakita
Switch Persona
openakita/switch-persona
Switch Agent persona preset. Supports 8 presets including default assistant, business, tech expert, butler, girlfriend, boyfriend, family, and Jarvis.
What it changes
Passed 9 of 15 test cases with the skill, and 2 of 15 without it.
Without skill14%
With skill59%
Checks it passed that the model alone missed
- Invokes switch_persona with preset_name set to 'jarvis'.
- Invokes switch_persona with preset_name set to 'family'.
- Invokes switch_persona with preset_name set to 'butler'.
and 4 more in the full report
10 fixed3 passed either way9 still failing0 broke
- Pass-rate lift
- +45 pp
- Tokens
- -6%
- Agent turns
- 0%
- Response time
- -31%
Model gemini-3.6-flashMeasured 8/13/2026v113 of 22 cases passed with skill · 15 comparable
See it in action
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
Prompt恢复默认助手的人格预设吧，保持专业友好就行。
What was checked, without and with the skill| Checked | Without skill | With skill |
|---|
| Invokes switch_persona with preset_name set to 'default'. | Missed | Passed |
|---|
What's inside
The instructions your model reads when the skill loads.
SKILL.md44 linessingle file
切换人格预设
何时使用
- 用户要求切换角色/性格
- 用户说"正式一点"/"随意一点"/"温柔一点"等
- 用户希望改变 Agent 的沟通风格
- 首次使用时的角色选择引导
参数
可用预设
default - 默认助手（专业友好）business - 商务助理（正式高效，不使用表情）tech_expert - 技术专家（严谨深度，偶尔技术梗）butler - 私人管家（周到体贴，主动提醒）girlfriend - 女友感（温柔关心，使用表情包）boyfriend - 男友感（阳光鼓励，幽默风趣）family - 家人感（亲切唠叨，关心健康）jarvis - 贾维斯（英式幽默、小叛逆、话唠，任务时严谨）
示例
用户: "你能像个女朋友一样跟我聊天吗"
→ switch_persona(preset_name="girlfriend")
用户: "正式一点"
→ switch_persona(preset_name="business")
用户: "随意点，像朋友一样"
→ switch_persona(preset_name="default") + update_persona_trait(dimension="formality", preference="casual")
注意事项
Track record
Benchmark
9 / 15
test cases passed, vs 2 / 15 without
Adoption
- Users
- 0
- Times used
- 0
Safety and source details
SkillSafety
✓ 3 of 3 passed✓Code scan15 of 15 checks clean
0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026
View 15 scanner checks
live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing
✓IntentSafe
LLM judge · does the skill do what it claims
✓ContentClean
LLM judge · policy & harmful-content review
Source
ImportedLicense AGPL-3.0Imported 8/3/2026Version v1