Repository navigation
v0.7.2 — Eval P1 fixes: prompt fidelity, multi-process safety, tool validation
opencode-plugin-loop 0.7.2
Patch release fixing the three P1 findings from the OpenCode vs Claude Code /loop comparison eval (L10/L11/L14), plus a small command addition.
Bug fixes
- LOOP-001 — prompt fidelity: scheduling flags (
--once,--all,--jitter=*) are now parsed only in the option prefix before the prompt begins, and--terminates flag parsing. The prompt is kept as the original substring, so flag-like text, whitespace, and newlines in the body are preserved verbatim (previously silently stripped — eval L14). - LOOP-002 — multi-process task safety: tasks record
ownerPid/ownerStartedAt. Store load drops only tasks whose owner process is confirmed dead (signal-0 probe, conservative on probe errors) instead of tombstoning everything written by a different live process. The instance lock now only serializes same-process plugin instances (shouldFire), so a second OpenCode process in the same project fires its own tasks instead of starving (eval L11). - LOOP-003 — tool entry validation: a shared
validateFixedInterval(≥ 1000ms) is now used by the slash parser,loop_schedulecreate, andset_fixed; the tool schema gains.min(1000); adaptive/maintenance create carryingintervalMsis rejected (eval L10 —loop_schedule(create, fixed, intervalMs=0)previously succeeded and silently fell back to 60s).
Added
- Bare
/loop stopcancels every task in scope (current session, or all sessions with--all)./loop stop <id>remains a single-task cancel alias.
Verification
npm test: 164/164 (new parser-fidelity, dual-process store,shouldFire, tool-validation, and stop tests)- Real OpenCode 1.18.5 E2E with two concurrent
opencode serveprocesses: L14 marker written verbatim with one-shot auto-cleanup, L11 dual-process coexistence + dead-owner cleanup, L10 tool rejection without ghost tasks
Full Changelog: b192f19...v0.7.2