Have Claude implement most of a day job's work
The poster gives Claude work tasks and has it implement most of the functionality, though the last 10% still takes the most time. Task: Suggested implementation: Build a .NET console app integrating Claude CLI and GitLab. The app polls assigned issues every 15 minutes, classifies readiness, drafts replies for unclear tickets, or triggers Claude to work on a new branch using repo context and attachments. Human role: Review generated code and replies before merging.; tools: Claude. Human role: The original post does not say how a person stayed involved.; limitations: Prerequisites: Access to Claude Max or equivalent high-quota API, GitLab repository permissions, and .NET environment. Constraint: Implement token budget checks to only start new tasks when >20% quota remains. Integration uncertainty: Handling image attachments and complex repo contexts may require custom parsing logic..
Tools
Published on Agent Work Atlas:
Overview
The poster gives Claude work tasks and has it implement most of the functionality, though the last 10% still takes the most time.
Workflow
Suggested implementation: Build a .NET console app integrating Claude CLI and GitLab. The app polls assigned issues every 15 minutes, classifies readiness, drafts replies for unclear tickets, or triggers Claude to work on a new branch using repo context and attachments. Human role: Review generated code and replies before merging.
Reported result
Outcome unknown due to missing source access. Proposed acceptance check: Measure reduction in manual coding time per ticket and verify that human review time stabilizes at 2-3 hours daily for a pilot batch of 10 issues.
Human role
The original post does not say how a person stayed involved.
Limitations
Prerequisites: Access to Claude Max or equivalent high-quota API, GitLab repository permissions, and .NET environment. Constraint: Implement token budget checks to only start new tasks when >20% quota remains. Integration uncertainty: Handling image attachments and complex repo contexts may require custom parsing logic.
Try it
Act as a DevOps engineer. I need to build a prototype automation loop for my team's GitLab issues using Claude CLI. Context: We have a .NET environment and a GitLab project [replaceable: project_url]. Inputs: Assigned issues with text and optional images. Output: A C# console application codebase and a README. Steps: 1. Create a .NET console app skeleton. 2. Implement a GitLab API client to fetch assigned issues every 15 minutes. 3. Add logic to classify issues: if unclear, draft a reply; if ready, trigger Claude CLI with the issue text and repo context. 4. Ensure Claude works on a new branch. 5. Add a safety check: only trigger new Claude tasks if API quota > 20%. 6. Output the code files and a dry-run simulation script. Quality Measure: The code must compile, and the simulation must correctly identify 3 test issues (1 unclear, 2 ready) without executing real API calls. Safeguards: Do not deploy to production. Run a local dry-run first. Ask me for the specific GitLab API token scope if needed.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.