---
id: d3f80a8b-24e9-407e-8b83-7b8886e16b08
type: procedure
title: "Bulk codebase audit and Gitea issue creation across multiple repos with parallel agents"
tags: [gitea, tea-cli, code-review, automation, major-domo, paper-dynasty, claude-code, agents]
importance: 0.7
confidence: 0.8
created: "2026-02-20T06:57:34.926191+00:00"
updated: "2026-02-20T06:57:34.926191+00:00"
---

# Bulk Codebase Audit + Gitea Issue Creation Pattern

## What Was Done
Conducted a full codebase audit across 5 repos (major-domo-v2, major-domo-database, paper-dynasty-discord, paper-dynasty-database, paper-dynasty-card-creation) using parallel Claude agents.

## Procedure

### Phase 1: Parallel Scan Agents
Launch 5 parallel Sonnet agents, one per repo. Each agent:
- `grep` for `TODO`, `FIXME`, `HACK`, `XXX`
- Reviews source files for: security issues, performance problems, error handling gaps, dead code, incomplete implementations
- Returns structured list of issues per repo

### Phase 2: Parallel Issue Creation Agents
Feed scan results to 5 parallel Bash agents that create Gitea issues:

```bash
tea issues create --repo cal/REPO --title "TITLE" --description "BODY"
```

## Key Learnings

1. `tea` CLI uses `--description` not `--body` for issue body text
2. `tea` labels require **numeric IDs**, not string names — embed label/category info inside the description instead
3. This scan-agents -> issue-creation-agents pattern parallelizes well across repos

## Results
- 112 total issues created: 21 + 20 + 25 + 26 + 20 across the 5 repos
- Pattern worked cleanly with no cross-agent interference