Warning
INTERNAL DEVELOPER DOCUMENT: This file is for project maintainers and contains technical debt audits, task backlogs, and roadmap tracking. For the official User Guide, please refer to the README.md.
Historical status note: This progress log is legacy context. It is not the current release contract. Use the frozen BETA_* docs in this directory for current beta truth.
- Daemon bootstrap & SQLite Ledger
- State machine & Transitions
- Basic Run/Step persistency
- Initial CLI & MCP layout
- Skeleton VS Code Extension
- Local Adapter Subprocess wrappers (Codex)
- Refactor & Strengthen Orchestration Runtime
- Make Codex Path Honest and Robust
- Artifact, Result, and Validation Retrieval
- Stronger Policy Model
- Recovery and Resumability
- MCP/Control Plane Completion
- VS Code Extension Completion
- Stronger Tests
- Formalize DSL Schemas for TaskSpec and ResultSpec
- Implement Validation Parsers for Explicit Payload Ingestion
- Plumb YAML Task Payload consumption directly into
orchestratorctlCLI - Plumb YAML Task Payload consumption directly into
orchestratorctlCLI - Scaffold Native Claude Code Adapter boundaries mapping anthropic capabilities
- Scaffold Native Qwen Adapter boundaries mapping local capabilities
- Deploy Universal Adapter Conformance Tests guaranteeing capability parity
- Implement Benchmark Metrics & SQLite Scoring Ledger
- Author
RoutingServicewith Smart Fallback Chaining - Scaffold Targeted IDE Chat Adapter for Extension-bound Proxies
- Deploy Compatibility Matrix Diagnostic Endpoints
- Complete E2E Validation of the Fully Operational Bridge
- Decompose
RunService.DispatchStepLifecycle - Fix
ToolRetryStepRun Identity Resolution - Implement Structured JSON Outputs for MCP Tools
- Harden Worktree Setup and Collision Handling
- Decouple Recovery Paths and Improve State Reconcile
- Inject Configuration-driven Paths into RunSvc (Self-Review)
- Implement Step Idempotency for MCP Retries (Self-Review)
- Phase V1.7: Improve log and result discoverability (V1.7 Complete)
- Strict Self-Review of Batch V1.5.2 Usability Hardening (Complete)
- Phase V1.8: Unified Configuration Engine (V1.x Roadmap)
- Modernize
ArtifactandValidationResultDomain Models - Hardened SQLite Schema for Detailed Evidence
- Implement Structured Retrieval in
RunService - Expose REST/MCP Endpoints for Validations
- Native CLI Commands for Inspection
- Verified missing-state robustness with unit tests
- Implement Exclusive Workspace Locking
- Add
RecoveryNotesto Run Ledger - Deep Reconciliation Engine for Stale Attempts
- Orphaned Worktree and Lock Cleanup
- Exposed Recovery Status to API/CLI/MCP
- Verified with Recovery Integration Tests
- Hierarchical Run -> Step -> [Gate, Artifact, Validation] Tree
- Status-aware Icons and Rich Tooltips
- Centralized API Client with Error Handling
- Actionable Commands: Approve/Reject Gate, Retry Step
- Structured Inspection: View Results/Validations in JSON buffers
- Corrected Backend Support for Step Retries
- Explicit Heuristic Routing (Deterministic Fallback)
- Simulation-Aware Benchmark Persistence
- REST API for Telemetry and Routing Configuration
- MCP Tools for Tactical Transparency
- Honest Documentation for Adapter Logic
- Configuration-driven Policy Engine (
PolicyRegistry) - Policy-bound Step Evaluation
- Robust API Integration Test Suite
- Audited CLI / API / MCP surface for functional completeness.
- Verified 100% test and build stability.
- Standardized all terminology (
State,lowercase_properties). - Produced Feature Status Matrix in
GAP_AUDIT.md. - Formalized Reviewer Guide in
README.md. - Qualified all project maturity claims with technical honesty.
- Audit existing task/input/result/state contracts (Micro-task complete)
- Define canonical planner-facing input contract (TaskSpec)
- Synchronize ResultSpec JSON schema with Go domain models
- Create canonical relay-facing result scaffold
- Opaque Streams:
stdout.logis captured but not yet streamed to the planner for real-time progress. - Reporting Detail: [RESOLVED] Results now include
RequestedAdapterand explicitIsSimulationflags for full transparency. - Clarify execution state semantics (timeout, needs_manual_attention)
- Formalize simulation semantics in output contract and documentation
- Audit execution state model (Micro-task complete)
- Define and document canonical execution/result state semantics
- Clarify lifecycle meaning of Runs, Steps, and Attempts
- Align manual-attention and simulation semantics
- Terminology uniformity (State vs Status)
- [-] Refactor Attempt state management (Moved to V1.1.3)
- [-] Consolidate intervention signaling (Moved to V1.1.3)
-
Audit existing CLI and identify gaps (Micro-task complete)
-
Align task submission with canonical contract (Micro-task complete)
-
Standardize machine-readable JSON output (Micro-task complete)
-
Refine CLI for reliable machine-readability (Batch V1.2.1 Complete)
-
Audit wait/result retrieval paths (Batch V1.2.2 Micro-task complete)
-
Align structured result retrieval CLI (Batch V1.2.2 Micro-task complete)
-
Add
waitsupport for terminal state monitoring (Batch V1.2.2 Complete) -
Refine CLI wait/result consistency (Batch V1.2.2 Complete)
-
Audit Codex adapter and identify hardening requirements (V1.3.1 Complete)
-
Validate adapter binary availability in
Start()(V1.3.1 Complete) -
Pass canonical task metadata to adapter command invocation (V1.3.1 Complete)
-
Implement Codex-specific result normalization (V1.3.1 Complete)
-
Align Codex adapter reporting with relay contracts (V1.3.1)
-
Audit Codex artifact harvesting flow (V1.3.2 Complete)
-
Harden Codex artifact discovery and metadata capture (V1.3.2 Complete)
-
Align Codex harvested outputs with canonical contracts (Batch V1.3.2 Complete)
-
Define and document local validation scenario (V1.3.2 Complete)
-
Add/align practical local validation path (V1.3.2 Complete)
-
Improve observable success/failure evidence (V1.3.2 Complete)
-
Add/strengthen validation coverage (V1.3.2 Complete)
-
Final alignment for validation scenario (Batch V1.3.3 Complete)
-
Audit execute-and-wait loop behavior (Phase V1.4.1 Complete)
-
Implement native timeout enforcement in RunService
-
Add/align practical timeout and polling controls (V1.4.3 Complete)
-
Align wait flow output with relay contract (V1.4.4 Complete)
-
Implement state discovery (run/step listing) (Batch V1.5.1 Complete)
-
Audit terminal outcome semantics (V1.4.5 Complete)
-
Clarify canonical terminal outcome meanings (V1.4.6 Complete)
-
Clarify manual-attention and retry reporting (V1.4.7 Complete)
-
Align terminal outcome semantics in CLI/Result output (V1.4.8 Complete)
-
Final alignment for Batch V1.4.2 (V1.4.9 Complete)
-
Audit local dev/config workflow (V1.5.5 Complete)
-
Align default paths and improve local startup (Batch V1.5.1 Complete)
-
Clarify execution modes (Real vs. Simulation) (Batch V1.5.1 Complete)
-
Implement unified config with env overrides (V1.x Roadmap)
-
Final alignment for Batch V1.5.1 (Micro-task Complete)
-
Audit maturity & limitations framing (V1.7.8 Complete)
-
Audit self-host/setup ergonomics (V1.7.9 Complete)
-
Harden self-host setup instructions (V1.8.0 Complete)
-
Harden self-host run flows (V1.8.1 Complete)
-
Practical publication-readiness cleanup (V1.8.2 Complete)
-
Audit repo for setup/start correctness (V1.8.3 Complete)
-
Fix command mismatches & align onboarding (V1.8.4 Complete)
-
Clarify config & execution modes (V1.8.5 Complete)
-
Consolidate one primary quickstart path (V1.8.6 Complete)
-
Final alignment for Phase V1.F1 (V1.8.7 Complete)
-
Phase V1.F2: Operational Maturity (V1.F2 Complete)
- Audit end-to-end workflow & identify friction (V1.F2.1)
- Implement
submit --waitand ID auto-generation (V1.F2.2) - Improve output inspection: results, artifacts, validations (V1.F2.3)
- Tighten failure/timeout/cancel recovery paths (V1.F2.4)
- Human-readable data rendering for all CLI lists (V1.F2.5)
-
[/] Phase V1.F3: Packaging & Governance (Active)
- Audit publication blockers and trust gaps (V1.F3.1 Complete)
- Resolve Critical Blockers (LICENSE, CONTRIBUTING, Makefile)
- Move/Ignore
codencer.db - Harden public-facing trust signals (V1.F3.2 Complete)
- Clarify public vs internal doc hierarchy (V1.F3.4 Complete)
- Implement
doctorbinary version checking - Author
CONTRIBUTING.md
-
Final alignment for Phase V1.C2 (Alignment Complete)
-
Implement
latestkeyword support for session memory (V1.x Roadmap) -
Implement incremental log streaming (
--follow) (V1.x Roadmap) -
Implement synthesized
step report(Outcome + Diffs + Validations) (V1.x Roadmap) -
Fix README Formatting & Align Quickstart Flow (Batch V1.L1 Complete)
-
Audit Operator Flow & Daily Ambiguities (Batch V1.L2 Complete)
-
Create canonical first-user local runbook (Batch V1.L2 Complete)
-
Improve CLI Clarity & Output Guidance (Batch V1.L2 Complete)
-
Audit Trust & Readiness Alignment (Batch V1.L3 Complete)
-
Finalize Trust & Readiness Alignment (Batch V1.L3 Complete)
-
Harden
doctorwith binary version checking (Batch V1.R1 Complete) -
Align Smoke Test with modern CLI ergonomics (Batch V1.R1 Complete)
-
Harden Primary Operator Sequence (submit -> wait -> inspect) (Batch V1.R1 Complete)
-
Tighten Result & Evidence Inspection Path (Visual Audit) (Batch V1.R1 Complete)
-
Clarify Non-Success Terminal Outcomes & Recovery (Batch V1.R1 Complete)
-
Fix EXAMPLES.md Rendering & Structure (Batch V1.R1 Complete)
-
Harden Canonical Operator Flow (README/EXAMPLES) (Batch V1.R1 Complete)
-
Final Readiness & Trust Alignment (v0.1.0-beta) (Batch V1.R1 Complete)
-
Batch R2: Final Metadata & Release Notes (Complete)
-
Implement "Latest" ID Alias for CLI (Batch V1.L3 Roadmap)
-
Setup Logic Automation (Batch V1.F6 Next)