Agent skill

fix-bug

Use when given a GitHub issue URL or number to investigate and implement a fix. Triggers on "fix issue", "fix bug", "fix

Stars 2,552
Forks 215

Install this agent skill to your Project

npx add-skill https://github.com/bytebase/dbhub/tree/main/.claude/skills/fix-bug

SKILL.md

Fix Bug from GitHub Issue

Systematic workflow for turning a GitHub issue into a working fix in the DBHub codebase.

Workflow

  1. Fetch → 2. Analyze → 3. Locate → 4. Reproduce → 5. Plan → 6. Implement → 7. Verify → 8. PR

Step 1: Fetch Issue

bash
# From URL: https://github.com/owner/repo/issues/123
gh issue view 123 --json title,body,labels,comments,state

# From another repo
gh issue view 123 --repo owner/repo --json title,body,labels,comments,state
Input How to fetch
https://github.com/owner/repo/issues/42 gh issue view 42 --repo owner/repo
#42 or 42 gh issue view 42 (current repo)
owner/repo#42 gh issue view 42 --repo owner/repo

Step 2: Analyze Issue

Extract from the issue:

  • What's broken: Expected vs actual behavior
  • Reproduction steps: How to trigger the bug
  • Environment: Database type, connection method (DSN, SSH tunnel, TOML config), transport (stdio/HTTP)
  • Labels/comments: May reveal affected area or prior investigation
  • Linked PRs/issues: Check for related context

Step 3: Locate Relevant Code

Use the issue details to identify which part of the codebase is affected. DBHub has a clear modular structure — most bugs fall into one of these areas:

Bug Category Where to Look Key Files
Connection failures Connector implementations src/connectors/{db-type}/index.ts, src/connectors/manager.ts
SQL execution errors Tool handlers src/tools/execute-sql.ts, src/utils/allowed-keywords.ts
Schema/table listing Search tool src/tools/search-objects.ts
DSN parsing issues Parser logic src/connectors/{db-type}/index.ts (DSNParser), src/utils/dsn-obfuscate.ts, src/utils/safe-url.ts
SSH tunnel problems Tunnel utilities src/utils/ssh-tunnel.ts, src/utils/ssh-config-parser.ts
TOML config issues Config loading src/config/toml-loader.ts, src/types/config.ts
Multi-database routing Manager & tools src/connectors/manager.ts, src/utils/tool-handler-helpers.ts
Custom tool issues Custom handler src/tools/custom-tool-handler.ts, src/tools/registry.ts
HTTP transport Server setup src/server.ts
Read-only violations SQL validation src/utils/allowed-keywords.ts, src/utils/sql-parser.ts
Row limiting SQL rewriting src/utils/sql-row-limiter.ts
API endpoint issues API handlers src/api/sources.ts, src/api/requests.ts
AWS IAM auth Token signing src/utils/aws-rds-signer.ts

Search for error messages, function names, or file paths mentioned in the issue. Trace the code path from entry point to the failure.

Step 4: Reproduce

If integration tests exist for the area: Write a failing test that captures the bug. DBHub's test infrastructure makes this straightforward:

  • Database connector bugs → extend existing integration test in src/connectors/__tests__/
  • Utility bugs → add cases to existing unit tests in src/utils/__tests__/
  • Tool handler bugs → add to src/tools/__tests__/
  • Config bugs → add to src/config/__tests__/

Use the test fixtures in src/__fixtures__/ for multi-database or readonly/max_rows scenarios.

If no test infrastructure applies: Trace the code path and confirm the logic flaw by reading.

Step 5: Plan the Fix

For non-trivial fixes (multi-file, architectural impact): use EnterPlanMode.

For simple fixes (single function, clear root cause): proceed directly.

Step 6: Implement

  • Fix the root cause, not just the symptom
  • Keep changes minimal and focused
  • Follow existing code conventions (see CLAUDE.md for style guide)
  • Use parameterized queries for any database operations
  • Validate inputs with zod schemas where appropriate

Step 7: Verify

Run the relevant tests to confirm the fix:

bash
pnpm test:unit                    # Quick check — no Docker needed
pnpm test src/path/to/test.ts     # Run the specific test file
pnpm test:integration             # Full integration suite if needed

Check that:

  • The failing test (if written) now passes
  • No existing tests regressed
  • The diff fully addresses the issue

Step 8: Create PR

  1. Create a branch:

    bash
    git checkout -b fix/issue-123
    
  2. Commit changes referencing the issue:

    bash
    git add <changed-files>
    git commit -m "$(cat <<'EOF'
    Fix: <short description>
    
    Closes #123
    EOF
    )"
    
  3. Push and create the PR:

    bash
    git push -u origin fix/issue-123
    gh pr create --title "Fix: <short description>" --body "$(cat <<'EOF'
    ## Summary
    <What was broken and how this fixes it>
    
    ## Changes
    <Bullet list of changes>
    
    Closes #123
    
    ## Test plan
    - [ ] Existing tests pass
    - [ ] New test covers the bug scenario (if applicable)
    EOF
    )"
    
  4. Return the PR URL to the user.

Common Mistakes

  • Fixing symptoms instead of root cause: Trace the full code path before patching
  • Skipping reproduction: A fix without a repro is a guess
  • Scope creep: Fix the reported issue, don't refactor surrounding code
  • Missing edge cases: Check if the fix handles related scenarios mentioned in comments
  • Not testing with the right database: If the bug is database-specific, test with that connector

Expand your agent's capabilities with these related and highly-rated skills.

bytebase/dbhub

testing

Run and troubleshoot tests for DBHub, including unit tests, integration tests with Testcontainers, and database-specific tests. Use when asked to run tests, fix test failures, debug integration tests, troubleshoot Docker/database container issues, or add new tests. Also use when verifying code changes work correctly or when CI test failures need investigation.

2,552 215
Explore
bytebase/dbhub

dbhub

Guide for querying databases through DBHub MCP server. Use this skill whenever you need to explore database schemas, inspect tables, or run SQL queries via DBHub's MCP tools (search_objects, execute_sql). Activates on any database query task, schema exploration, data retrieval, or SQL execution through MCP — even if the user just says "check the database" or "find me some data." This skill ensures you follow the correct explore-first workflow instead of guessing table structures.

2,552 215
Explore
davila7/claude-code-templates

verl-rl-training

Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

23,776 2,298
Explore
davila7/claude-code-templates

openrlhf-training

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

23,776 2,298
Explore
davila7/claude-code-templates

gguf-quantization

GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.

23,776 2,298
Explore
davila7/claude-code-templates

Claude Code Guide

Master guide for using Claude Code effectively. Includes configuration templates, prompting strategies "Thinking" keywords, debugging techniques, and best practices for interacting with the agent.

23,776 2,298
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results