hq

The Problem: Multiple AI agents stepping on each other

# You spawn 4 Claude agents to work on a project.
# Agent 1 starts editing file A.
# Agent 2 also starts editing file A.
# Agent 3 depends on Agent 1's output but starts early.
# Agent 4 crashes silently  - its work is lost.
#
# No coordination = wasted work, conflicts, and confusion.

Solution

$ # Create project with dependency-tracked steps
hq create_project --name "refactor-auth" --steps '[
  {"name": "update-models", "deps": []},
  {"name": "update-handlers", "deps": ["update-models"]},
  {"name": "update-tests", "deps": ["update-handlers"]},
  {"name": "update-docs", "deps": []}
]'

Output

{"project_id":"proj_abc123","name":"refactor-auth","steps":[{"id":"step_1","name":"update-models","status":"pending","deps":[]},{"id":"step_2","name":"update-handlers","status":"blocked","deps":["update-models"]},{"id":"step_3","name":"update-tests","status":"blocked","deps":["update-handlers"]},{"id":"step_4","name":"update-docs","status":"pending","deps":[]}]}