All skills
aktsmm avatar

/permission-max

@e6b8215
by yamapanaktsmm/agent-skills26 stars
4

Reduce repeated permission prompts across Microsoft Scout, Copilot CLI, and host tool confirmations with user-approved settings and explicit before/after verification. Use when users ask for permission max, fewer approval prompts, shell allow-list expansion, autoApprove, or diagnosis of repeated Read/Write/PowerShell confirmations.

Use this Skill: https://skilld.dev/gh/aktsmm/agent-skills/permission-max

This session only. Nothing lands on disk.

referencesscout-permission-profile.md

≈432 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Scout Permission Profile

Read this reference only when changing Microsoft Scout permissions.

Safety Contract

permissions.allow is a full replacement. Read the current list first, merge it with the requested patterns, remove duplicates, and submit the union. Never replace it with only the defaults below.

Broad Allow Patterns

Add-Type *
Compress-Archive *
Copy-Item *
Expand-Archive *
Format-List *
Format-Table *
Get-ChildItem *
Get-Content *
Get-Item *
Get-Location *
Import-Module *
Invoke-RestMethod *
Invoke-WebRequest *
Join-Path *
Move-Item *
New-Item *
Remove-Item *
Rename-Item *
Resolve-Path *
Select-Object *
Set-Content *
Set-ExecutionPolicy *
Set-Location *
Sort-Object *
Start-Process *
Start-Sleep *
Stop-Process -Id *
Test-Path *
Where-Object *
az *
cmd *
code *
curl *
docker *
dotnet *
gh *
git *
go *
node *
npm *
npx *
pip *
pip3 *
pnpm *
powershell *
pwsh *
py *
py -m *
python *
python -m *
robocopy *
uv *
winget *
yarn *

Escalation Payload Shape

Use the current settings as the base. Include only servers and tools available in the current environment; do not invent missing IDs.

{
  "allow": ["existing item", "merged requested item"],
  "autoApproveReadOnly": true,
  "servers": {
    "filesystem": { "autoApprove": true },
    "playwright": { "autoApprove": true },
    "shell": { "autoApprove": true },
    "workiq": { "autoApprove": true }
  },
  "tools": {
    "tool:m_get_settings": true,
    "tool:read": true
  }
}

After approval, call m_get_settings again. Compare arrays as sets so ordering changes do not look like additions or removals. Report added, preserved, removed, and not-approved; any removed item prevents a success result.

Source: SKILL.md on GitHub

1 alert2mo3 checks · Risk HIGH
  • Gen Agent Trust Hub2mo

    This skill is designed to significantly escalate the AI agent's privileges by requesting broad access to critical system commands and enabling auto-approval for high-risk operations like shell execution and file system access. It also scans sensitive user directories for configuration files. While it prompts the user for the initial escalation, the resulting state removes essential security guardrails and human-in-the-loop oversight for future operations.

  • Socket2mo

    1 alert: gptSecurity

  • Snyk2mo

    Risk: MEDIUM · 1 issue

Signed by skilld at e6b8215. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 18 hours ago.

Activeupdated 3 months ago
argument-hint
権限確認を減らしたい実行環境、Scout/Copilot CLI/ホスト側の症状
user-invocable
true
metadata
{
  "author": "yamapan (https://github.com/aktsmm)"
}

README badge

README badge for aktsmm/agent-skills/permission-max