Skip to content

v0.11.0

Choose a tag to compare

View all tags
@clintecker clintecker released this 27 Mar 03:47
v0.11.0
ed7c1e3

Autopilot Mode

Run pipelines fully autonomous with LLM-backed human gate decisions.

# Balanced judgment (default)
tracker --autopilot mid build_product

# Move fast, approve everything reasonable
tracker --autopilot lax build_product

# High quality bar — pushes back on gaps
tracker --autopilot hard build_product

# Approves but writes detailed feedback for later review
tracker --autopilot mentor build_product

# Deterministic auto-approve — no LLM, for CI/testing
tracker --auto-approve build_product

Personas

Persona Bias Style
lax Strong approve "Looks fine, moving on"
mid Balanced Concise engineering judgment
hard Prefers retry/adjust Cites specific quality gaps
mentor Approve with notes 3-5 sentences of constructive feedback

Design

  • Uses the pipeline's existing LLM client (no extra config)
  • Structured JSON output: {"choice": "approve", "reasoning": "..."}
  • Fallback to default edge on LLM error (with warning)
  • One retry before fallback
  • All decisions logged for post-hoc review

Also in this release

  • dippin-lang upgraded to v0.10.0

🤖 Generated with Claude Code

Assets 11
Loading