simplicio-loop v3.38.0 — Multi-agent coordination

simplicio-loop v3.38.0 — Multi-agent coordination

This release is about one hard problem that only shows up once several agent sessions work the
same repo at once: how does a session know what's already claimed, what's merged-but-incomplete,
and what to do with its own idle time instead of duplicating a sibling's work?

Highlights

• scripts/coordinator.py — the decision core. Given live GitHub state (claim comments + merged
  PRs), returns one deterministic action per issue: OWN, CONTINUE_OWN, DEFER_ACTIVE_CLAIM,
  RECLAIM_STALE, or VERIFY_PARTIAL, plus a duplicate_risk flag. Caught, live, a real collision:
  two sessions independently building a findings collector for the same issue under two different
  filenames.
• scripts/pr_dod_review.py — the reviewer for idle time. When every open issue is claimed,
  reviews open PRs against the 7-dimension Definition of Done and the underlying issue's frozen
  acceptance-criteria checklist instead of idling. Verified against a real merged "MVP slice" PR:
  correctly flagged 17 of 17 acceptance criteria on the parent epic as still unresolved.
• scripts/finding_collector.py (#466, phase 1) — durable, fingerprinted, deduplicated defect
  memory: the same underlying bug collapses into one record no matter how many agents/runs observe
  it.
• scripts/release_check.py — compares the local canonical version against the latest GitHub
  release and tells the driving LLM to update instead of silently working on a stale checkout.
  Fail-open when offline; wired into scripts/doctor.py.
• Two new documented conventions: references/multi-agent-coordination.md and
  references/background-verification.md, wired into SKILL.md's triage step.
• scripts/evolution.py, scripts/workflow_topology.py, scripts/agent_replication.py — first MVP
  slices of the Continuous Evolution (#467), Adaptive Architecture (#468), and Elastic Replication
  (#469) epics.
• scripts/worktree_cleanup.py (#484) — mandatory post-merge branch/worktree cleanup.
• WI-471 CLI additions (preflight subcommand, --json on status) and WI-466 findings
  completion-gate wiring.
• Two live regressions caught and fixed on main itself this cycle — a PR that silently deleted a
  function definition, and a subsequent squash-merge race that reintroduced the exact same bug.
  Both found by actually running the affected script, not by trusting a green PR description.
• Full README overhaul (English + all 14 translations): removed stale "roadmap" framing for work
  that's actually shipped (#422–436), fixed stale skill/runtime counts, refreshed the flow diagram,
  added a Star History section.
• Test suite grew to 231 files (2,544 collected tests); scripts/claims_audit.py stayed at 14/14.

Usage

pip install simplicio-loop
simplicio-loop install            # current project
simplicio-loop install --global   # user-wide

Then:

/simplicio-loop finish all the open issues

See CHANGELOG.md for
the full list.