permission-tuner skill
Analyze permission prompts and auto mode denials, then generate permission rules and autoMode entries that cut prompt fatigue. Use when prompts keep interrupting, auto mode falls back to prompting, or after sessions with many denials.
Is the permission-tuner skill safe?
Clean: nothing in its files matched our rules. We read 1 file in the folder on 2026-09-28.
No findings.
Install the permission-tuner skill
A skill is a folder. Copy it into your agent's skills folder and the agent loads it when the task matches its description.
git clone --depth 1 https://github.com/rohitg00/pro-workflow.git /tmp/pro-workflow mkdir -p ~/.claude/skills cp -r /tmp/pro-workflow/skills/permission-tuner ~/.claude/skills/permission-tuner
In the Claude apps, zip the folder and upload it from the Skills settings. The folder on GitHub
The instructions your agent would load
SKILL.md as published, without the frontmatter. Read it on GitHub
Permission Tuner
Reduce permission prompts by fixing their causes, not by piling up rules.
Trigger
Use when:
- Permission prompts interrupt flow repeatedly
- Auto mode keeps blocking routine work or falls back to prompting everything
- Starting a new project and want to configure permissions
How Prompts Happen
- Rules resolve first: deny > ask > allow. A matching ask rule prompts even in auto mode, and a PreToolUse hook returning allow cannot override it.
- In auto mode, anything no rule resolves goes to a classifier. After 3 blocks in a row or 20 in a session, auto mode pauses and prompts for everything until you approve one.
- Auto mode drops broad allow rules: Bash(), wildcarded interpreters like Bash(node ), package-manager run commands, Agent, and Monitor.
- Hooks that exit 2 on harmless commands count as denials and push the session toward that fallback.
Workflow
Step 1: Gather data
Rules merge across scopes, so check every file that exists, not only the user file. /permissions shows the rules Claude Code actually loaded, and claude auto-mode config shows the effective autoMode (the classifier ignores autoMode in .claude/settings.local.json).
for f in ~/.claude/settings.json .claude/settings.json .claude/settings.local.json; do
[ -f "$f" ] && jq --arg f "$f" '{file: $f, mode: .permissions.defaultMode, allow: (.permissions.allow|length), ask: (.permissions.ask|length), deny: (.permissions.deny|length), autoMode: (.autoMode != null)}' "$f"
done
claude auto-mode config
ls -t ~/.claude/projects/*/*.jsonl | head -20 | xargs grep -ho '"toolDenialKind":"[^"]*"' | sort | uniq -c | sort -rntoolDenialKind values seen in transcripts (undocumented, may change): permission-rule (rule or hook), automode-blocked (classifier), automode-parsing-error and automode-unavailable (no verdict, not tunable), user-rejected (you said no).
Step 2: Pick the right fix
Step 3: Generate rules
{
"permissions": {
"allow": [
"Bash(git status)",
"Bash(npm test)",
"Bash(npm run lint)"
],
"ask": [
"Bash(git push --force*)",
"Bash(git push * main)",
"Bash(git push * main *)",
"Bash(git push *:main)",
"Bash(git push *:main *)",
"Bash(git push * master)",
"Bash(git push * master *)",
"Bash(git push *:master)",
"Bash(git push *:master *)",
"Bash(gh pr merge *)",
"Bash(npm publish*)"
],
"deny": [
"Bash(rm -rf *)",
"Read(**/.env)",
"Read(**/.env.*)"
]
},
"autoMode": {
"environment": [
"$defaults",
"Source control: repositories under github.com/your-org are trusted; feature-branch pushes are routine."
],
"allow": ["$defaults"],
"soft_deny": [
"$defaults",
"Merging a pull request needs the user's approval for that specific merge."
]
}
}autoMode only takes effect in ~/.claude/settings.json, managed settings, or --settings. Keep "$defaults" in each list. Check the result with claude auto-mode config and review custom rules with claude auto-mode critique.
Output
PERMISSION TUNER REPORT
Mode: [mode] Rules: [X] allow, [Y] ask, [Z] deny autoMode: [yes/no]
Denials: [n] permission-rule, [n] classifier, [n] no verdict, [n] you rejected
Remove (ask rules you always approve):
- Bash(...) -- approved [n]x
Add to autoMode.environment:
+ "..." -- clears [n] blocks
Add to autoMode.allow / soft_deny:
+ "..."
Fix hooks:
! [script] blocked [n] harmless commands
Keep asking:
~ force push, merge, publish, deploy
Estimated prompts saved per session: ~[N]Rules
- Prefer autoMode entries over new ask rules
- Destructive operations stay in deny or ask
- Always present changes for user approval before applying
- Never write autoMode into project settings
- Include estimated prompt savings
More skills from rohitg00/pro-workflow
- Aagent-teamsCoordinate multiple Claude Code sessions as a team — lead + teammates with shared task lists, mailbox messaging, and file-lock claiming. Patterns for team sizing, task decomposition, and when to use teams vs sub-agents vs worktrees.
- Aauto-setupAuto-configure quality gates, hooks, and settings for a new project. Detects project type and sets up appropriate tooling. Use when onboarding a new codebase.
- Abatch-orchestrationDecompose large-scale changes into independent units and spawn parallel agents in isolated worktrees. Use for migrations, refactors, codemods, and any change touching 10+ files with the same pattern.
- Cbug-captureCapture a user-reported defect as a durable GitHub issue written in the project's own domain language. Explores the codebase in parallel for context but never leaks file paths or line numbers into the issue. Use when the user reports a bug conversationally, runs a QA pass, or says "file an issue", "log this as a bug", "capture this".
- Acompact-guardSmart context compaction with state preservation. Saves critical files, task progress, and working state before compaction, restores after. Use before manual compact or when auto-compact triggers.
- Acontext-engineeringMaster the four operations of context engineering — Write, Select, Compress, Isolate. Manage token budgets, compaction strategies, and context partitioning to keep AI sessions sharp and efficient.
- Acontext-optimizerOptimize token usage and context management. Use when sessions feel slow, context is degraded, or you're running out of budget.
- Acost-trackerTrack session costs, set budget alerts, and optimize token spend. Use to check costs mid-session or set spending limits.
- Adesign-engineeringApply interface craft when building or reviewing UI - motion, easing, timing, springs, component feel, and visual foundations. Use when building a component, animation, transition, hover or press state, modal, drawer, toast, or when polishing an interface so it feels right. Says "make this feel better", "add an animation", "polish the UI", "review this component".
- AdeslopRemove AI-generated code slop, unnecessary comments, and over-engineering from the current branch diff. Cleans up boilerplate, simplifies abstractions, strips defensive code, and in skill-file mode lints SKILL.md files for quality. Use when cleaning up code, simplifying, removing boilerplate, before committing, or when reviewing a skill before promoting it.
- Adomain-modelingBuild the project's shared language and bounded contexts before writing code, so names stay consistent and the agent stops paraphrasing domain concepts. Produces a CONTEXT.md glossary and decision records. Use at the start of a project or feature, or when the codebase and the people describing it speak different languages.
- Afile-watcherConfigure file watching hooks to auto-react to config changes, env file updates, and dependency modifications. Use to set up reactive workflows.