Skip to content

Document AgentSync gitignore modes and team workflows step by step #274

@yacosta738

Description

@yacosta738

Summary

We are keeping [gitignore].enabled = true as the default, but the product needs much better documentation explaining exactly how both workflows work:

  1. Default workflow: AgentSync manages .gitignore and collaborators recreate symlinks via agentsync apply / package-manager hooks
  2. Opt-out workflow: teams disable gitignore management and commit symlinks intentionally

Right now the behavior exists, but the trade-offs and team instructions are not documented clearly enough.

Goal

Produce end-to-end docs that explain the full decision tree and repo lifecycle step by step.

Required documentation topics

  • What .agents/ is and why it is the canonical source of truth
  • What [gitignore].enabled = true does by default
  • Why this is the current default
  • What collaborators must do after cloning in the default workflow
  • How to wire agentsync apply into prepare / install flows
  • How to opt out by setting [gitignore].enabled = false
  • How to intentionally commit symlinks instead
  • Recommended git staging/commit flow after migration
  • Common pitfalls and how to recover from them
  • Cross-links from getting-started, configuration reference, and CLI docs

Acceptance criteria

  • Documentation covers both workflows with explicit trade-offs
  • Docs include concrete command examples and a recommended step-by-step migration flow
  • Docs explain collaborator expectations for each mode
  • Docs remain consistent with the current product default (enabled = true)
  • At least getting-started, configuration reference, and CLI docs are updated or cross-linked appropriately

Notes

This subissue is intentionally documentation-heavy. The goal is to leave this area extremely clear for users and teammates.

Metadata

Metadata

Assignees

Type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions