Skip to content

CHANGELOG.md

Latest commit

 

History

History
146 lines (85 loc) · 6.99 KB

CHANGELOG.md

File metadata and controls

146 lines (85 loc) · 6.99 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project follows SemVer where practical.

[Unreleased]

Added

Changed

Fixed

[0.1.10] - 2026-03-09

Added

  • Added roboppi agents tasks supersede and roboppi agents status get|set for file-backed workflow status summaries and task replacement flow.
  • Extended the agents workflow DSL with seed task id/depends_on/tags/requires_plan_approval, member role, and worker defaultArgs.
  • Added TUI Timeline/Raw/Usage tabs with structured LLM log summaries, agent runtime usage stats, and dormant-agent visibility.

Changed

  • Agent coordination now prunes stopped resident agents before reconcile, skips dormant members, restores mailbox body previews, and emits task-superseded activity events.
  • The Codex CLI adapter now normalizes legacy sandbox/approval flags and merges task/profile defaultArgs before dispatch.

Fixed

  • Clarified agent task claimability and dependency handling, including seeded dependency resolution and broader test coverage.
  • Added verification around supervised IPC capability detection, TUI state/runtime accounting, and prompt-byte tracking in worker costs.

[0.1.9] - 2026-03-06

Added

  • roboppi agents chat interactive REPL, message reply, and resident-agent dispatch loops so teammates can keep working across the full workflow run.
  • Runner-owned lead inbox summaries at _agents/inbox-summary.json, dynamic membership reconciliation, and new example workflows for live team reconfiguration.
  • TUI agent overview/chat surfaces for inline messaging and richer agent activity inspection.

Changed

  • Renamed the swarm-facing CLI, docs, examples, tests, and runtime modules to the agents surface so the feature matches the shipped vocabulary.
  • Agent coordination now keeps long-lived LLM teammates resident and routes lead communication through mailbox-backed workflow services instead of prompt-level polling.

Fixed

  • Added reply-token handling and expanded unit/acceptance coverage around agent mailbox, broker, and TUI chat behavior.

[0.1.8] - 2026-03-02

Added

  • Agents v1: file-backed mailbox + task store for agent teams under <context_dir>/_agents/.
  • New roboppi agents CLI (JSON-only stdout) for members, messaging, tasks, and housekeeping (incl. message recv --wait-ms, claim-token ack, stale recovery).
  • Workflow YAML agents: config for defining members + seeding tasks; inject ROBOPPI_AGENTS_* env vars into steps.
  • Workflow TUI Agents tab + agent_* exec events for metadata-only activity visibility.
  • New worker capabilities MAILBOX and TASKS to gate Agents tooling (Claude Code allowlist support).
  • Runner-owned LeadInboxBroker for auto-consuming lead inbox and maintaining _agents/inbox-summary.json.
  • Dynamic membership via roboppi agents members set|upsert|remove with coordinator reconcile loop.

Changed

  • Agents conformance checklist is captured in docs/spec/agents.md and enforced by the self-dev Agents impl loop.

Fixed

  • Harden roboppi agents tool-facing inputs with explicit path-safety validation and JSON-safe error output.

[0.1.7] - 2026-03-01

Added

  • Workflow Management Agent: optional supervisor hooks (pre_step, pre_check, post_check, on_stall) that can annotate, skip, adjust timeouts, or abort workflows based on step state.
  • Management hook event isolation: management worker events are written to per-invocation JSONL under context/_management/ and do not leak into the main workflow telemetry/TUI stream.
  • Example self-dev workflow for iterating on the management-agent implementation review (examples/workflow-management-agent-impl-review-loop.yaml) + bootstrap helper (scripts/self-dev/workflow-management-agent-impl-review/bootstrap.sh).

Changed

  • Workflow parsing for steps.<id>.management is stricter: unknown keys are rejected, and step-level per-hook overrides are supported.
  • Claude Code adapter uses --output-format stream-json automatically for streaming tasks when configured for JSON output.
  • Worker tasks accept optional variant hints (plumbed through to adapters that support it).
  • Embedded Pi SDK management engine is read-only in v1 (no command execution/edit/write tools) to preserve mechanism/policy separation.

Fixed

  • Branch verification no longer leaks ambient BASE_BRANCH-related environment into isolated workflow runs.

[0.1.6] - 2026-02-25

Added

  • Supervised workflow TUI now streams worker events to 2: Logs in real time (stdout/stderr/progress/patch) via IPC job_event messages.

Changed

  • In supervised + TUI mode, stdout/stderr forwarding to 2: Logs is enabled by default; set ROBOPPI_TUI_STREAM_STDIO=0 to forward progress/patch only.

Docs

  • Updated workflow/TUI docs and quickstarts.

[0.1.5] - 2026-02-25

Added

  • Allow subworkflow steps (workflow:) to use completion_check loops (max_iterations) to rerun child workflows until complete.
  • Add subworkflow event bubbling (bubble_subworkflow_events, subworkflow_event_prefix) and deterministic export copying (exports_mode: merge|replace).
  • Improve workflow TUI: ANSI-aware wrapping in tabs, and y to copy the visible detail tab to the clipboard (system clipboard commands or OSC52 fallback).
  • Add a runnable subworkflow loop example (examples/subworkflow-loop.yaml, examples/subworkflow-loop-child.yaml).

Changed

  • Completion checks accept decisions from decision_file JSON or text markers (PASS/FAIL/COMPLETE/INCOMPLETE), and can fall back to parsing worker stdout (JSON or marker).
  • Completion check instructions interpolate $ROBOPPI_COMPLETION_CHECK_ID / ${ROBOPPI_COMPLETION_CHECK_ID} to the active check id for robustness.
  • scripts/agent-pr-loop/todo-check.sh now respects ROBOPPI_TODO_PATH / ROBOPPI_LOOP_DIR and prefers legacy .agentcore-loop when present.

[0.1.4] - 2026-02-22

Fixed

  • Stabilize OpenTUI workflow TUI rendering by sanitizing ANSI control sequences and using ANSI-aware width/truncation.

[0.1.3] - 2026-02-22

Added

  • Subworkflow steps via workflow:: parent workflows can invoke child workflows, with isolated child context under _subworkflows/ and optional exports to copy child artifacts into the parent context.
  • Recursion protection for subworkflow invocations (cycle detection + max nesting depth).
  • CLI E2E coverage for subworkflow execution in three modes: direct, supervised (stdio transport), supervised (socket transport).

Changed

  • Agent catalog loading is consolidated in src/workflow/workflow-loader.ts and reused by CLI and daemon; child workflows inherit the parent's resolved catalog.
  • make test / bun run test are scoped to this repo's test suites (test/unit, test/integration, tests/at) to avoid running vendored/reference tests.

Fixed

  • Harden workflow parsing and context path handling against path traversal and reserved-name collisions (step IDs and artifact names).

[0.1.2] - 2026-02-20

  • See GitHub release notes.