v0.2.1

What's Changed

• Improve README overview and usage by @fazpu in #51
• Prepare 0.2.1 release by @fazpu in #52

Full Changelog: v0.2.0...v0.2.1

v0.2.0

What's Changed

• feat: initial http coordinator/worker loop with team-harness by @fazpu in #1
• feat: simplify API to two endpoints, drop leases and polling by @fazpu in #2
• 0.0.1 by @fazpu in #3
• chore: add harness tooling config by @fazpu in #4
• chore: remove feature planning and outputs by @fazpu in #5
• chore: ignore _additional_context directory by @fazpu in #7
• docs: clarify root config file path in README by @fazpu in #6
• chore!: bump team-harness to 0.2.0 and rename Harness SDK to TeamHarness by @fazpu in #8
• docs: fix missed team_harness.Harness reference in README by @fazpu in #9
• feat!: prefix team-harness config fields with team_harness_ by @fazpu in #10
• docs: add atomic_bomberman_lol example scaffold by @fazpu in #11
• feat!: expose team_harness_provider + reject unknown config fields by @fazpu in #12
• feat: skip api-key env-var check when provider does not need it by @fazpu in #13
• docs: add loopy-loop skill by @fazpu in #14
• chore: update Makefile codex tooling by @fazpu in #15
• feat: derive goal hash from goal by @fazpu in #16
• chore: ignore agent directories by @fazpu in #17
• chore: ignore Claude prompts by @fazpu in #18
• Finish podcast creator reusable loop example by @fazpu in #19
• Extend podcast workflow prompts by @fazpu in #20
• Install eval-banana dev skill by @fazpu in #21
• Update loopy-loop skill for recent workflow changes by @fazpu in #22
• Configure team-harness worker models from loopy-loop by @fazpu in #23
• Add session user updates and harness artifacts by @fazpu in #24
• Persist harness error details by @fazpu in #25
• Add inner_outer_eval init template by @fazpu in #26
• Update inner_outer_eval eval reviewer prompt by @fazpu in #27
• Prefix session ids with timestamp by @fazpu in #28
• Update loopy session state and delivery policies by @fazpu in #29
• Use session control as loop stop switch by @fazpu in #30
• Nudge workflow prompts toward upfront research by @fazpu in #31
• Bias outer workflow toward integrated implementation by @fazpu in #33
• Treat suspected blockers as follow-up tasks by @fazpu in #34
• Add inner workflow skill capture guidance by @fazpu in #35
• chore: switch claude make targets to --remote-control by @fazpu in #32
• Restrict eval workflows to harness judge checks by @fazpu in #36
• chore: bump inner_outer_eval template defaults by @fazpu in #37
• Retry failed workflow when scheduling is stuck by @fazpu in #38
• Require team-harness 0.2.8 by @fazpu in #39
• Expose coordinator retry controls by @fazpu in #40
• Recover completed loopy tasks after worker handoff failure by @fazpu in #41
• Implement workflow sets and child sessions by @fazpu in #42
• Add PM planner dispatcher template by @fazpu in #43
• Require team-harness 0.2.10 by @fazpu in #44
• WRI-547 Update template agent model defaults by @fazpu in #45
• Remove atomic bomberman template by @fazpu in #46
• Upgrade gpt-5.4 references to gpt-5.5 by @fazpu in #47
• Fix project URLs by @fazpu in #48
• Prepare package releases by @fazpu in #49
• Bump version to 0.2.0 by @fazpu in #50

New Contributors

• @fazpu made their first contribution in #1

Full Changelog: https://github.com/writeitai/loopy-loop/commits/v0.2.0