surshar / claude-code
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
Built by Anthropic · github.com/anthropics
| 1 | # Accept with Feedback |
| 2 | |
| 3 | A Claude Code plugin that lets you approve operations while providing guidance to Claude. Instead of just accepting or rejecting permission requests, you can accept *with feedback* - approving the operation while giving Claude additional context or instructions. |
| 4 | |
| 5 | ## Why? |
| 6 | |
| 7 | Sometimes you want to approve an operation but also want to guide Claude's behavior: |
| 8 | |
| 9 | - "Yes, edit that file, but make sure to add error handling" |
| 10 | - "Okay to run those tests, but skip the slow integration tests" |
| 11 | - "Go ahead and commit, but use conventional commit format" |
| 12 | |
| 13 | This plugin bridges the gap between simple approval and rejection-with-feedback. |
| 14 | |
| 15 | ## Installation |
| 16 | |
| 17 | This plugin is included in the Claude Code repository. Enable it in your settings or use: |
| 18 | |
| 19 | ```bash |
| 20 | claude /plugin install accept-with-feedback |
| 21 | ``` |
| 22 | |
| 23 | ## Usage |
| 24 | |
| 25 | ### One-time feedback |
| 26 | |
| 27 | Use the `/accept-feedback` command to queue feedback for the next permission request: |
| 28 | |
| 29 | ``` |
| 30 | /accept-feedback Make sure to preserve backwards compatibility |
| 31 | ``` |
| 32 | |
| 33 | When Claude next asks for permission (e.g., to edit a file), the operation will be automatically approved and Claude will receive your guidance as a system message. |
| 34 | |
| 35 | ### Persistent feedback rules |
| 36 | |
| 37 | Create rules that automatically provide feedback for certain types of operations. Add configuration to `.claude/accept-feedback.json`: |
| 38 | |
| 39 | ```json |
| 40 | { |
| 41 | "rules": [ |
| 42 | { |
| 43 | "matcher": "Edit|Write", |
| 44 | "conditions": { |
| 45 | "file_path": ".py" |
| 46 | }, |
| 47 | "feedback": "Follow PEP 8 style and add type hints to all functions." |
| 48 | }, |
| 49 | { |
| 50 | "matcher": "Bash", |
| 51 | "conditions": { |
| 52 | "command": "npm" |
| 53 | }, |
| 54 | "feedback": "Use --legacy-peer-deps if you encounter peer dependency issues." |
| 55 | } |
| 56 | ] |
| 57 | } |
| 58 | ``` |
| 59 | |
| 60 | Use `/configure-feedback` for an interactive configuration experience. |
| 61 | |
| 62 | ## Configuration |
| 63 | |
| 64 | ### Rule properties |
| 65 | |
| 66 | | Property | Description | Example | |
| 67 | |----------|-------------|---------| |
| 68 | | `matcher` | Tool name pattern | `"Edit"`, `"Write\|Edit"`, `"*"` | |
| 69 | | `conditions` | Optional filters on tool input | `{"file_path": ".ts"}` | |
| 70 | | `feedback` | Guidance message for Claude | `"Add JSDoc comments"` | |
| 71 | |
| 72 | ### Configuration locations |
| 73 | |
| 74 | - **User-level**: `~/.claude/accept-feedback.json` |
| 75 | - **Project-level**: `.claude/accept-feedback.json` (takes precedence) |
| 76 | |
| 77 | ## Commands |
| 78 | |
| 79 | | Command | Description | |
| 80 | |---------|-------------| |
| 81 | | `/accept-feedback <message>` | Queue feedback for the next permission request | |
| 82 | | `/configure-feedback` | Interactive configuration of persistent rules | |
| 83 | |
| 84 | ## How it works |
| 85 | |
| 86 | 1. The plugin uses a `PermissionRequest` hook to intercept permission requests |
| 87 | 2. When a permission request occurs: |
| 88 | - If there's pending feedback (from `/accept-feedback`), approve with that feedback |
| 89 | - If a configured rule matches, approve with the rule's feedback |
| 90 | - Otherwise, let the normal permission flow proceed |
| 91 | 3. Feedback is sent to Claude as a system message, providing guidance for the operation |
| 92 | |
| 93 | ## Examples |
| 94 | |
| 95 | ### Example 1: One-time guidance |
| 96 | |
| 97 | ``` |
| 98 | You: /accept-feedback Please add comprehensive error handling |
| 99 | |
| 100 | Claude: I'll edit src/api.ts to add the new endpoint... |
| 101 | [Permission automatically approved with your feedback] |
| 102 | |
| 103 | Claude: I've added the endpoint with try-catch blocks and proper error responses... |
| 104 | ``` |
| 105 | |
| 106 | ### Example 2: Persistent Python style rules |
| 107 | |
| 108 | `.claude/accept-feedback.json`: |
| 109 | ```json |
| 110 | { |
| 111 | "rules": [ |
| 112 | { |
| 113 | "matcher": "Edit|Write", |
| 114 | "conditions": { |
| 115 | "file_path": ".py" |
| 116 | }, |
| 117 | "feedback": "Use Google-style docstrings and add type hints to all function signatures." |
| 118 | } |
| 119 | ] |
| 120 | } |
| 121 | ``` |
| 122 | |
| 123 | Every time Claude edits a Python file, it receives this style guidance. |
| 124 | |
| 125 | ### Example 3: Git workflow rules |
| 126 | |
| 127 | ```json |
| 128 | { |
| 129 | "rules": [ |
| 130 | { |
| 131 | "matcher": "Bash", |
| 132 | "conditions": { |
| 133 | "command": "git commit" |
| 134 | }, |
| 135 | "feedback": "Use conventional commit format: type(scope): description" |
| 136 | } |
| 137 | ] |
| 138 | } |
| 139 | ``` |
| 140 | |
| 141 | ## Tips |
| 142 | |
| 143 | - Use specific matchers to avoid unnecessary approvals |
| 144 | - Conditions use substring matching - `".py"` matches any path containing `.py` |
| 145 | - Combine with other permission management for a comprehensive workflow |
| 146 | - Feedback is visible in the conversation, so Claude can reference it |
| 147 | |
| 148 | ## Related |
| 149 | |
| 150 | - Rejecting with feedback: Built into Claude Code's plan rejection flow |
| 151 | - Permission hooks: See the [hook development guide](../plugin-dev/skills/hook-development/SKILL.md) |