- Home
- Skills
- Testing & Debugging
- Software Issue Solver
Software Issue Solver
Investigate software issues, identify root causes, implement fixes, run tests, and prepare review-ready changes—with human approval at every consequential step. Works with the development tools available to your agent.
Free
Works with the AI Agents you already use
Software Issue Solver
Example session with this skill installed
Analyze the bug reported in issue #402 regarding the login timeout and propose a fix. The repo is in the current directory.
- Read your context and instructions
- Compiled the software issue solver
I have investigated issue #402 regarding the login timeout. The issue appears to be caused by an incorrect millisecond conversion in auth-provider.ts.
Issue Analysis and Proposed Fix
| Component | Status | Details |
|---|---|---|
| Current Branch | main | Clean working tree detected. |
| Hypothesis | Logic Error | SESSION_TIMEOUT is treated as seconds but passed to a function expecting ms. |
| Affected Files | src/auth/auth-provider.ts | Line 142 needs a * 1000 multiplier. |
| Verification | Unit Test | Run npm test src/auth/auth-provider.test.ts to confirm fix. |
| Risk | Low | Localized change to session logic; no schema changes required. |
Proposed Plan
- Create a new branch
fix/issue-402-login-timeout. - Update the timeout calculation in
auth-provider.ts. - Execute the auth unit test suite.
- Prepare a draft Pull Request for your review.
Next steps
- Please type "approve" to begin the implementation on a new branch.
- Provide any additional test cases you want me to include in the verification.
- Specify if you prefer a different branch naming convention.
Connects securely to your tools. The creator never sees your data.
About this skill
Software Issue Solver
Turn software issues into tested, review-ready solutions.
Software Issue Solver is a human-in-the-loop development skill that helps an AI coding agent investigate bugs, understand feature requests, inspect a repository, implement fixes, and verify the results.
What it does
- Understand: Read the issue, acceptance criteria, comments, and reproduction steps.
- Investigate: Inspect repository instructions, relevant code, tests, dependencies, and existing changes.
- Plan: Identify a likely cause, propose a solution, and explain risks.
- Get approval: Ask the developer to approve the implementation plan before changing code.
- Implement: Make focused changes while preserving unrelated work.
- Verify: Run relevant tests and report actual results.
- Prepare for review: Create a draft pull request, merge request, or equivalent handoff when supported.
Platform independent
The workflow does not assume GitHub, GitLab, or a particular issue tracker. It adapts to the tools, repository host, and permissions available in the agent's environment.
Human control comes first
The skill does not independently approve, merge, deploy, release, or close issues. Final consequential decisions remain with the developer.
Who is it for?
Developers, maintainers, solo builders, and engineering teams who want an AI coding agent to do more than suggest a fix: investigate the issue, implement the change, verify it, and prepare it for human review.
The goal: Less time spent on repetitive issue investigation, with clear evidence and developer control throughout.
Cannot independently merge, deploy, or close issues. Requires manual setup of repository credentials and tool access in the execution environment.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 13 days ago
- Free with an Agensi account
Permissions
Reads repository instructions and relevant source, test, and configuration files. Writes only files needed for the approved issue. Preserves unrelated changes and avoids secrets, credentials, and environment files. Actual access is limited by the agent's existing permissions.