Home

How the AI Agent Pipeline Works

GitLab issues and comments are the entire interface. There is no separate dashboard or CLI: every action begins with an issue, a comment, or a merge request.

Trigger

  1. HumanGitLab: open issue (type::feature / type::bug / type::chore)
  2. GitLabOrchestrator: Issue webhook event
  3. OrchestratorOrchestrator: run pipeline (background)
  4. OrchestratorGitLab: open merge request

Retry

  1. HumanGitLab: comment "/retry" on the issue
  2. GitLabOrchestrator: Note webhook event
  3. OrchestratorOrchestrator: re-run pipeline, reuse existing branch
  4. OrchestratorGitLab: update existing MR

Fix

  1. HumanGitLab: comment "/fix <feedback>" on the MR
  2. GitLabOrchestrator: Note webhook event
  3. OrchestratorOrchestrator: resolve issue via "Closes #N" in MR description
  4. OrchestratorOrchestrator: seed feedback as a failed review verdict
  5. OrchestratorGitLab: update existing MR

Merge & close

  1. HumanGitLab: merge the MR
  2. GitLabOrchestrator: Merge request webhook event
  3. OrchestratorGitLab: set workflow::done, prune merged branch locally

The Pipeline

Drag any node to inspect the execution graph. It will return to its resting position when released.

verdict passedretry budget leftcritical failure OR retries exhausted

Issue opened

or /retry · /fix comment

Prepare

clone target repo,

checkout issue branch

Planner Agent

LLM breaks the issue

into ordered sub-tasks

Development Agent

delegates to OpenCode —

the only agent that writes code

Review Agent

automated checks first (fast, free),

LLM judgment only if those pass

Supervisor

deterministic — no LLM

Git Agent

commit, push,

open/update MR

Blocked —

flagged for human review

Merge request ready

Agents never talk to each other directly. Every agent reads from and writes to one shared state object (graph/state.py); the orchestrator — not the agents — decides what happens next.