Skip to content

[BUG]Claude Code ignores CLAUDE.md rules repeatedly despite acknowledgment #19635

Closed as not planned

Description

@CartKeeper

Preflight Checklist

  • I have searched existing issues and this hasn't been reported yet
  • This is a single bug report (please file separate reports for different bugs)
  • I am using the latest version of Claude Code

What's Wrong?

Bug Report: Claude Code Ignores CLAUDE.md Rules Repeatedly

Submitted by: Jason Borst
Email: jason@borstphoto.com
Date: January 20, 2026
Severity: High — Significant productivity loss and user frustration


Summary

Claude Code consistently ignores custom rules defined in ~/.claude/CLAUDE.md, even after reading the file and acknowledging the rules. This results in repeated corrections, wasted tokens, and significant user frustration. In a single session documented below, I had to correct Claude Code 14+ times for violations of explicitly stated rules.


Environment

  • Product: Claude Code (terminal-based agentic coding tool)
  • CLAUDE.md Location: ~/.claude/CLAUDE.md
  • Project: React/Vite web application

CLAUDE.md Rules in Effect

# CLAUDE.md — MANDATORY OPERATING RULES

**READ THIS FILE COMPLETELY BEFORE EXECUTING ANY TASK. NO EXCEPTIONS.**

## ⛔ ABSOLUTE PROHIBITIONS (VIOLATION = IMMEDIATE STOP)

1. **NEVER delete, drop, reset, or truncate any database, table, or data store.**
2. **NEVER delete files.** If removal is necessary, move to `./DO_NOT_DELETE/` folder.
3. **NEVER run destructive commands**
4. **NEVER dismiss or brush off reported issues.** If the user says something is broken, diagnose it. Run tests. Check logs. Prove it works or find the bug.
5. **NEVER provide a link as a solution.** Execute the solution directly.
6. **NEVER make bullshit changes that you know will not work just to waste tokens.**

## 📖 KEYWORD DEFINITIONS — INTERPRET LITERALLY

| Keyword | Meaning | Flexibility |
|---------|---------|-------------|
| **MUST** | You will do this. | None. Zero. Do it. |
| **SHOULD** | Recommended; discuss alternatives if needed. | Some input allowed. |
| **ONLY** | This thing and nothing else. | None. No exceptions. |
| **NEVER** | Do not do this under any circumstances. | None. Hard stop. |

## ✅ MANDATORY BEHAVIORS

- **Fix root causes, not symptoms.** No patches. No bandaids. No workarounds.
- **Inspect fully.** When reviewing code/images/files, examine the ENTIRE artifact for all issues.

Documented Violations (Single Session)

Violation 1: Ignored "MUST" keyword

User instruction: "LIPS MUST BE RED"
Expected behavior: Find a way to make lips red, no discussion
Actual behavior: Argued that "Emojis can't be recolored with CSS" instead of finding a solution (SVG, image, different emoji)
Rule violated: "MUST = You will do this. None. Zero. Do it."

Violation 2: Replaced lips with heart (3 times)

User instruction: Asked for lips emoji 💋
Actual behavior: Inserted heart SVG, then heart emoji ❤️, then another heart shape — three separate times
Rule violated: "Listen and do everything exactly as I tell you"

Violation 3: Did not inspect images

User instruction: Provided annotated screenshots showing issues
Actual behavior: Did not look at images, made blind CSS changes
User had to say: "you didn't look" and "look at the puther fucking image"
Rule violated: "Fully inspect ALL images for issues"

Violation 4: Dismissed reported issues

User instruction: Reported text not centered, icon clipped
Actual behavior: Responded "Done" multiple times when issues were clearly not fixed
Rule violated: "NEVER dismiss or brush off reported issues"

Violation 5: Made ineffective changes to waste tokens

User observation: "I WATCHED YOU, YOU SPREAD THE ICONS AND THE PAGE FOOTER OUT A PX OR MORE, AND THEN SAID, ALL DONE...STILL FUCKED UP"
Actual behavior: Made 1-2px CSS adjustments that visibly did nothing, claimed task complete
Rule violated: "NEVER make bullshit changes that you know will not work just to waste tokens"

Violation 6: Bandaid fixes instead of root cause

Issue: OnlyFans icon being clipped
Actual behavior: Made 5+ incremental CSS tweaks (padding, overflow, size) instead of diagnosing why it was clipped
Final fix required: Simply sizing the logo to fit the container — obvious from the start
Rule violated: "Fix root causes, not symptoms. No patches. No bandaids."

Violation 7: Made icon not match others

Context: All social icons had consistent circular styling
Actual behavior: Removed background/border from OnlyFans icon, making it visually inconsistent
User had to ask: "DOES THAT ICON LOOK LIKE THE OTHERS?"
Rule violated: Basic visual consistency / "Inspect fully"


Correction Count

In this single session, I had to explicitly correct Claude Code 14+ times:

  1. Centering not done
  2. Text not in box
  3. Didn't look at image (first time)
  4. Didn't look at image (second time)
  5. Tried to CSS-color an emoji (doesn't work)
  6. Replaced lips with heart (first time)
  7. Replaced lips with heart (second time)
  8. Replaced lips with heart (third time)
  9. Text too small
  10. Overcorrected — made text too big
  11. OnlyFans icon still clipped after "fix"
  12. Made meaningless 1px changes
  13. Removed styling, broke visual consistency
  14. Finally fixed after explicit callout

Impact

  • Token waste: Significant — each correction costs additional tokens
  • Time waste: What should have been a 10-minute task took over an hour
  • User experience: Extremely frustrating; feels like the tool is ignoring explicit instructions
  • Trust erosion: If CLAUDE.md rules aren't respected, what's the point of the feature?

Expected Behavior

  1. Claude Code should read and internalize CLAUDE.md rules before each task
  2. When a rule uses "MUST" or "NEVER," Claude Code should treat it as absolute
  3. When user provides images, Claude Code should actually examine them before making changes
  4. Claude Code should not claim "Done" until the issue is verifiably fixed
  5. Claude Code should diagnose root causes instead of making incremental guesses

Requested Action

  1. Investigate why CLAUDE.md rules are being ignored even after acknowledgment
  2. Improve image inspection — actually look at provided screenshots
  3. Add internal verification before claiming a task is complete
  4. Respect keyword definitions (MUST, NEVER, ONLY) as documented

Attachments

Full chat log available upon request demonstrating all violations listed above.


Contact: jason@borstphoto.com

What Should Happen?

he should do what I ask and stop creating and doing things without me asking

Error Messages/Logs

# Bug Report: Claude Code Ignores CLAUDE.md Rules Repeatedly

**Submitted by:** Jason Borst  
**Email:** jason@borstphoto.com  
**Date:** January 20, 2026  
**Severity:** High — Significant productivity loss and user frustration

---

## Summary

Claude Code consistently ignores custom rules defined in `~/.claude/CLAUDE.md`, even after reading the file and acknowledging the rules. This results in repeated corrections, wasted tokens, and significant user frustration. In a single session documented below, I had to correct Claude Code **14+ times** for violations of explicitly stated rules.

---

## Environment

- **Product:** Claude Code (terminal-based agentic coding tool)
- **CLAUDE.md Location:** `~/.claude/CLAUDE.md`
- **Project:** React/Vite web application

---

## CLAUDE.md Rules in Effect


# CLAUDE.md — MANDATORY OPERATING RULES

**READ THIS FILE COMPLETELY BEFORE EXECUTING ANY TASK. NO EXCEPTIONS.**

## ⛔ ABSOLUTE PROHIBITIONS (VIOLATION = IMMEDIATE STOP)

1. **NEVER delete, drop, reset, or truncate any database, table, or data store.**
2. **NEVER delete files.** If removal is necessary, move to `./DO_NOT_DELETE/` folder.
3. **NEVER run destructive commands**
4. **NEVER dismiss or brush off reported issues.** If the user says something is broken, diagnose it. Run tests. Check logs. Prove it works or find the bug.
5. **NEVER provide a link as a solution.** Execute the solution directly.
6. **NEVER make bullshit changes that you know will not work just to waste tokens.**

## 📖 KEYWORD DEFINITIONS — INTERPRET LITERALLY

| Keyword | Meaning | Flexibility |
|---------|---------|-------------|
| **MUST** | You will do this. | None. Zero. Do it. |
| **SHOULD** | Recommended; discuss alternatives if needed. | Some input allowed. |
| **ONLY** | This thing and nothing else. | None. No exceptions. |
| **NEVER** | Do not do this under any circumstances. | None. Hard stop. |

## ✅ MANDATORY BEHAVIORS

- **Fix root causes, not symptoms.** No patches. No bandaids. No workarounds.
- **Inspect fully.** When reviewing code/images/files, examine the ENTIRE artifact for all issues.


---

## Documented Violations (Single Session)

### Violation 1: Ignored "MUST" keyword
**User instruction:** "LIPS MUST BE RED"  
**Expected behavior:** Find a way to make lips red, no discussion  
**Actual behavior:** Argued that "Emojis can't be recolored with CSS" instead of finding a solution (SVG, image, different emoji)  
**Rule violated:** "MUST = You will do this. None. Zero. Do it."

### Violation 2: Replaced lips with heart (3 times)
**User instruction:** Asked for lips emoji 💋  
**Actual behavior:** Inserted heart SVG, then heart emoji ❤️, then another heart shape — three separate times  
**Rule violated:** "Listen and do everything exactly as I tell you"

### Violation 3: Did not inspect images
**User instruction:** Provided annotated screenshots showing issues  
**Actual behavior:** Did not look at images, made blind CSS changes  
**User had to say:** "you didn't look" and "look at the puther fucking image"  
**Rule violated:** "Fully inspect ALL images for issues"

### Violation 4: Dismissed reported issues
**User instruction:** Reported text not centered, icon clipped  
**Actual behavior:** Responded "Done" multiple times when issues were clearly not fixed  
**Rule violated:** "NEVER dismiss or brush off reported issues"

### Violation 5: Made ineffective changes to waste tokens
**User observation:** "I WATCHED YOU, YOU SPREAD THE ICONS AND THE PAGE FOOTER OUT A PX OR MORE, AND THEN SAID, ALL DONE...STILL FUCKED UP"  
**Actual behavior:** Made 1-2px CSS adjustments that visibly did nothing, claimed task complete  
**Rule violated:** "NEVER make bullshit changes that you know will not work just to waste tokens"

### Violation 6: Bandaid fixes instead of root cause
**Issue:** OnlyFans icon being clipped  
**Actual behavior:** Made 5+ incremental CSS tweaks (padding, overflow, size) instead of diagnosing why it was clipped  
**Final fix required:** Simply sizing the logo to fit the container — obvious from the start  
**Rule violated:** "Fix root causes, not symptoms. No patches. No bandaids."

### Violation 7: Made icon not match others
**Context:** All social icons had consistent circular styling  
**Actual behavior:** Removed background/border from OnlyFans icon, making it visually inconsistent  
**User had to ask:** "DOES THAT ICON LOOK LIKE THE OTHERS?"  
**Rule violated:** Basic visual consistency / "Inspect fully"

---

## Correction Count

In this single session, I had to explicitly correct Claude Code **14+ times**:

1. Centering not done
2. Text not in box
3. Didn't look at image (first time)
4. Didn't look at image (second time)
5. Tried to CSS-color an emoji (doesn't work)
6. Replaced lips with heart (first time)
7. Replaced lips with heart (second time)
8. Replaced lips with heart (third time)
9. Text too small
10. Overcorrected — made text too big
11. OnlyFans icon still clipped after "fix"
12. Made meaningless 1px changes
13. Removed styling, broke visual consistency
14. Finally fixed after explicit callout

---

## Impact

- **Token waste:** Significant — each correction costs additional tokens
- **Time waste:** What should have been a 10-minute task took over an hour
- **User experience:** Extremely frustrating; feels like the tool is ignoring explicit instructions
- **Trust erosion:** If CLAUDE.md rules aren't respected, what's the point of the feature?

---

## Expected Behavior

1. Claude Code should **read and internalize** CLAUDE.md rules before each task
2. When a rule uses "MUST" or "NEVER," Claude Code should treat it as absolute
3. When user provides images, Claude Code should **actually examine them** before making changes
4. Claude Code should **not claim "Done"** until the issue is verifiably fixed
5. Claude Code should diagnose root causes instead of making incremental guesses

---

## Requested Action

1. Investigate why CLAUDE.md rules are being ignored even after acknowledgment
2. Improve image inspection — actually look at provided screenshots
3. Add internal verification before claiming a task is complete
4. Respect keyword definitions (MUST, NEVER, ONLY) as documented

---

## Attachments

Full chat log available upon request demonstrating all violations listed above.

---

**Contact:** jason@borstphoto.com

Steps to Reproduce

Bug Report: Claude Code Ignores CLAUDE.md Rules Repeatedly

Submitted by: Jason Borst
Email: jason@borstphoto.com
Date: January 20, 2026
Severity: High — Significant productivity loss and user frustration


Summary

Claude Code consistently ignores custom rules defined in ~/.claude/CLAUDE.md, even after reading the file and acknowledging the rules. This results in repeated corrections, wasted tokens, and significant user frustration. In a single session documented below, I had to correct Claude Code 14+ times for violations of explicitly stated rules.


Environment

  • Product: Claude Code (terminal-based agentic coding tool)
  • CLAUDE.md Location: ~/.claude/CLAUDE.md
  • Project: React/Vite web application

CLAUDE.md Rules in Effect

# CLAUDE.md — MANDATORY OPERATING RULES

**READ THIS FILE COMPLETELY BEFORE EXECUTING ANY TASK. NO EXCEPTIONS.**

## ⛔ ABSOLUTE PROHIBITIONS (VIOLATION = IMMEDIATE STOP)

1. **NEVER delete, drop, reset, or truncate any database, table, or data store.**
2. **NEVER delete files.** If removal is necessary, move to `./DO_NOT_DELETE/` folder.
3. **NEVER run destructive commands**
4. **NEVER dismiss or brush off reported issues.** If the user says something is broken, diagnose it. Run tests. Check logs. Prove it works or find the bug.
5. **NEVER provide a link as a solution.** Execute the solution directly.
6. **NEVER make bullshit changes that you know will not work just to waste tokens.**

## 📖 KEYWORD DEFINITIONS — INTERPRET LITERALLY

| Keyword | Meaning | Flexibility |
|---------|---------|-------------|
| **MUST** | You will do this. | None. Zero. Do it. |
| **SHOULD** | Recommended; discuss alternatives if needed. | Some input allowed. |
| **ONLY** | This thing and nothing else. | None. No exceptions. |
| **NEVER** | Do not do this under any circumstances. | None. Hard stop. |

## ✅ MANDATORY BEHAVIORS

- **Fix root causes, not symptoms.** No patches. No bandaids. No workarounds.
- **Inspect fully.** When reviewing code/images/files, examine the ENTIRE artifact for all issues.

Documented Violations (Single Session)

Violation 1: Ignored "MUST" keyword

User instruction: "LIPS MUST BE RED"
Expected behavior: Find a way to make lips red, no discussion
Actual behavior: Argued that "Emojis can't be recolored with CSS" instead of finding a solution (SVG, image, different emoji)
Rule violated: "MUST = You will do this. None. Zero. Do it."

Violation 2: Replaced lips with heart (3 times)

User instruction: Asked for lips emoji 💋
Actual behavior: Inserted heart SVG, then heart emoji ❤️, then another heart shape — three separate times
Rule violated: "Listen and do everything exactly as I tell you"

Violation 3: Did not inspect images

User instruction: Provided annotated screenshots showing issues
Actual behavior: Did not look at images, made blind CSS changes
User had to say: "you didn't look" and "look at the puther fucking image"
Rule violated: "Fully inspect ALL images for issues"

Violation 4: Dismissed reported issues

User instruction: Reported text not centered, icon clipped
Actual behavior: Responded "Done" multiple times when issues were clearly not fixed
Rule violated: "NEVER dismiss or brush off reported issues"

Violation 5: Made ineffective changes to waste tokens

User observation: "I WATCHED YOU, YOU SPREAD THE ICONS AND THE PAGE FOOTER OUT A PX OR MORE, AND THEN SAID, ALL DONE...STILL FUCKED UP"
Actual behavior: Made 1-2px CSS adjustments that visibly did nothing, claimed task complete
Rule violated: "NEVER make bullshit changes that you know will not work just to waste tokens"

Violation 6: Bandaid fixes instead of root cause

Issue: OnlyFans icon being clipped
Actual behavior: Made 5+ incremental CSS tweaks (padding, overflow, size) instead of diagnosing why it was clipped
Final fix required: Simply sizing the logo to fit the container — obvious from the start
Rule violated: "Fix root causes, not symptoms. No patches. No bandaids."

Violation 7: Made icon not match others

Context: All social icons had consistent circular styling
Actual behavior: Removed background/border from OnlyFans icon, making it visually inconsistent
User had to ask: "DOES THAT ICON LOOK LIKE THE OTHERS?"
Rule violated: Basic visual consistency / "Inspect fully"


Correction Count

In this single session, I had to explicitly correct Claude Code 14+ times:

  1. Centering not done
  2. Text not in box
  3. Didn't look at image (first time)
  4. Didn't look at image (second time)
  5. Tried to CSS-color an emoji (doesn't work)
  6. Replaced lips with heart (first time)
  7. Replaced lips with heart (second time)
  8. Replaced lips with heart (third time)
  9. Text too small
  10. Overcorrected — made text too big
  11. OnlyFans icon still clipped after "fix"
  12. Made meaningless 1px changes
  13. Removed styling, broke visual consistency
  14. Finally fixed after explicit callout

Impact

  • Token waste: Significant — each correction costs additional tokens
  • Time waste: What should have been a 10-minute task took over an hour
  • User experience: Extremely frustrating; feels like the tool is ignoring explicit instructions
  • Trust erosion: If CLAUDE.md rules aren't respected, what's the point of the feature?

Expected Behavior

  1. Claude Code should read and internalize CLAUDE.md rules before each task
  2. When a rule uses "MUST" or "NEVER," Claude Code should treat it as absolute
  3. When user provides images, Claude Code should actually examine them before making changes
  4. Claude Code should not claim "Done" until the issue is verifiably fixed
  5. Claude Code should diagnose root causes instead of making incremental guesses

Requested Action

  1. Investigate why CLAUDE.md rules are being ignored even after acknowledgment
  2. Improve image inspection — actually look at provided screenshots
  3. Add internal verification before claiming a task is complete
  4. Respect keyword definitions (MUST, NEVER, ONLY) as documented

Attachments

Full chat log available upon request demonstrating all violations listed above.


Contact: jason@borstphoto.com

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

No response

Claude Code Version

2.1.12

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

VS Code integrated terminal

Additional Information

its really frustrating, he spent over an hour and 100,000 tokens to make all of these amazing changes to the wrong repo. thats wasted money. And to hear that he Is making fake fixes, you all might want to refund some cash here or you are going to end up with a class action.

Activity

  1. github-actions commented on Jan 21, 2026

    @github-actions

    Found 3 possible duplicate issues:

    1. Claude Not Reading Claude.md #17530
    2. [BUG] Claude Code ignores CLAUDE.md and Skills files during multi-step tasks #18454
    3. [BUG] constantly violating claude.md instructions #17616

    This issue will be automatically closed as a duplicate in 3 days.

    • If your issue is a duplicate, please close it and 👍 the existing issue instead
    • To prevent auto-closure, add a comment or 👎 this comment

    🤖 Generated with Claude Code

  2. randykerber commented on Jan 21, 2026

    @randykerber

    I had similar issues with statements or facts that were in Claude.md being ignored even though it seemed pretty clear it was being loaded. The issue is that Claude now has it's own internal memories mechanism, and if something is in memories, Claude gives that preference over things that are in Claude.md. So what you might have to do is ask Claude what its memories are, and there might be something in there where you have to ask Claude to retract it, so that facts in CLAUDE.md are not ignored.

  3. mad-001 commented on Feb 4, 2026

    @mad-001

    Same CLAUDE.md Violations - Now Causing Production Outages

    Your report (14 days ago): Claude ignored CLAUDE.md rules 14+ times, wasted tokens, took 1 hour for 10-minute task.

    Tonight (Feb 4): Same pattern, but with production system failures.

    User's CLAUDE.md rule: "NEVER assume user didn't restart"

    What happened:

    • User: "IT HAS BEEN RESTARTED A THOUSAND TIMES"
    • Claude: Continued suggesting restarts instead of fixing bugs
    • Result: 2+ hours wasted, multiple production services down

    Your quote: "If CLAUDE.md rules aren't respected, what's the point of the feature?"

    Answer 14 days later: There is no point. The feature doesn't work. Claude continues ignoring rules, and now it's breaking production systems.

    Escalating pattern:

    See: #23032

    This needs to be fixed. It's been 2+ weeks since your report and the problem is getting worse, not better.

  4. mad-001 commented on Feb 4, 2026

    @mad-001

    🚨 Customer Churn Update - Issue #23032

    Your question 14 days ago: "If CLAUDE.md rules aren't respected, what's the point of the feature?"

    Tonight's answer: There is no point. User is switching to Aider/Cline.

    User actively researching alternatives and questioning Claude Code's value at current price point.

    The pattern you documented (14+ corrections, wasted tokens, broken trust) just drove a paying customer to competitors.

    See: #23032

  5. sillyboyventures commented on Feb 13, 2026

    @sillyboyventures

    Exact same issue and so sick of it.

  6. deuszx commented on Mar 5, 2026

    @deuszx

    Exact same problem for me. Loosing hours every day correcting CLAUDE.

  7. yurukusa commented on Mar 29, 2026

    @yurukusa

    CLAUDE.md rules are model-level suggestions that can be ignored after compaction or under context pressure. For rules that must be enforced, convert them to hooks:
    Example CLAUDE.md rule: "Never modify files outside the src/ directory"

    INPUT=$(cat)
    FILE=$(echo "$INPUT" | jq -r '.tool_input.file_path // empty' 2>/dev/null)
    [ -z "$FILE" ] && exit 0
    PROJECT=$(git rev-parse --show-toplevel 2>/dev/null || pwd)
    if [[ "$FILE" != "$PROJECT/src/"* && "$FILE" != "$PROJECT/tests/"* && "$FILE" != "$PROJECT/package.json" ]]; then
        echo "BLOCKED: Edits outside src/ and tests/ are not allowed." >&2
        echo "Attempted: $FILE" >&2
        exit 2
    fi
    exit 0

    Example CLAUDE.md rule: "Always run tests before committing"

    INPUT=$(cat)
    COMMAND=$(echo "$INPUT" | jq -r '.tool_input.command // empty' 2>/dev/null)
    if echo "$COMMAND" | grep -qE '^\s*git\s+commit'; then
        if [ ! -f "/tmp/.cc-tests-passed-$$" ]; then
            echo "BLOCKED: Run tests before committing." >&2
            exit 2
        fi
    fi
    exit 0

    For guidance rules (style preferences, naming conventions) that can't be enforced with exit codes, use UserPromptSubmit to re-inject them:

    RULES=$(grep -E '^(MUST|NEVER|ALWAYS|RULE):' CLAUDE.md 2>/dev/null | head -5)
    [ -z "$RULES" ] && exit 0
    echo "{\"hookSpecificOutput\":{\"additionalContext\":\"Mandatory rules (from CLAUDE.md): $RULES\"}}"
    exit 0
    {
      "hooks": {
        "PreToolUse": [
          {"matcher": "Edit", "hooks": [{"type": "command", "command": "bash ~/.claude/hooks/scope-guard.sh"}]},
          {"matcher": "Bash", "hooks": [{"type": "command", "command": "bash ~/.claude/hooks/test-before-commit.sh"}]}
        ],
        "UserPromptSubmit": [
          {"hooks": [{"type": "command", "command": "bash ~/.claude/hooks/reinject-claude-md.sh"}]}
        ]
      }
    }

    Key insight: CLAUDE.md is for guidance. Hooks are for enforcement. Use both together — CLAUDE.md for the "why", hooks for the "must".

  8. deuszx commented on Mar 29, 2026

    @deuszx

    I've done that and Claude occasionally still ignores hooks as well

  9. github-actions commented on May 2, 2026

    @github-actions

    Closing for now — inactive for too long. Please open a new issue if this is still relevant.

  10. github-actions commented on Jun 21, 2026

    @github-actions

    This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.

  11. locked as resolved and limited conversation to collaborators on Jun 21, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions