feat(corrections): implement append mode for CorrectionService — add guidance without recomputing #9735

Open
opened 2026-04-15 07:46:37 +00:00 by HAL9000 · 1 comment
Owner

Metadata

  • Commit message: feat(corrections): implement append mode for CorrectionService — add guidance without recomputing
  • Branch name: feat/v3.3.0-correction-service-append-mode

Background and Context

Epic #4959 (Correction Engine) requires an append mode for plan correction. Unlike revert mode, append mode adds new guidance to the plan without invalidating or recomputing existing decisions. This allows users to steer the plan in a new direction without losing prior work.

This issue is part of the v3.3.0 milestone (M4: Corrections + Subplans + Checkpoints) and blocks Epic #4959.

Expected Behavior

CorrectionService.correct(plan_id, decision_id, mode="append", guidance=...) adds guidance to the plan without invalidating downstream decisions. The guidance is stored in the correction_attempts table and the plan continues execution from the current state with the new guidance applied. The plan.attempt counter increments after each append correction.

Acceptance Criteria

  • CorrectionService.correct(plan_id, decision_id, mode="append", guidance=...) is implemented
  • Append mode adds guidance to the plan without invalidating downstream decisions
  • The guidance is stored in the correction_attempts table with mode="append"
  • The plan continues execution from the current state with the new guidance applied
  • plan.attempt counter increments after each append correction
  • BDD/Gherkin tests cover append mode correction flow
  • Unit tests achieve >= 97% coverage

Subtasks

  • Implement CorrectionService.correct() append mode branch
  • Implement guidance storage in correction_attempts table
  • Implement plan continuation with new guidance applied
  • Write BDD/Gherkin feature file for correction append flow
  • Write unit tests for append mode
  • Verify nox passes with coverage >= 97%

Definition of Done

  1. Append mode is implemented and adds guidance without recomputing
  2. Guidance is persisted to correction_attempts table
  3. plan.attempt increments after each append correction
  4. BDD tests pass and coverage >= 97%
  5. Code reviewed and merged to main branch
  • Parent Epic: #4959 (EPIC: Correction Engine — Revert & Append Modes)
  • Parent Legendary: #4942 (LEGENDARY: Decision Intelligence)
  • Blocks: #4959
  • Depends on: Issue 1 (CorrectionService foundation)

Automated by CleverAgents Bot
Agent: new-issue-creator

## Metadata - **Commit message**: `feat(corrections): implement append mode for CorrectionService — add guidance without recomputing` - **Branch name**: `feat/v3.3.0-correction-service-append-mode` ## Background and Context Epic #4959 (Correction Engine) requires an append mode for plan correction. Unlike revert mode, append mode adds new guidance to the plan without invalidating or recomputing existing decisions. This allows users to steer the plan in a new direction without losing prior work. This issue is part of the v3.3.0 milestone (M4: Corrections + Subplans + Checkpoints) and blocks Epic #4959. ## Expected Behavior `CorrectionService.correct(plan_id, decision_id, mode="append", guidance=...)` adds guidance to the plan without invalidating downstream decisions. The guidance is stored in the `correction_attempts` table and the plan continues execution from the current state with the new guidance applied. The `plan.attempt` counter increments after each append correction. ## Acceptance Criteria - [ ] `CorrectionService.correct(plan_id, decision_id, mode="append", guidance=...)` is implemented - [ ] Append mode adds guidance to the plan without invalidating downstream decisions - [ ] The guidance is stored in the `correction_attempts` table with `mode="append"` - [ ] The plan continues execution from the current state with the new guidance applied - [ ] `plan.attempt` counter increments after each append correction - [ ] BDD/Gherkin tests cover append mode correction flow - [ ] Unit tests achieve >= 97% coverage ## Subtasks - [ ] Implement `CorrectionService.correct()` append mode branch - [ ] Implement guidance storage in `correction_attempts` table - [ ] Implement plan continuation with new guidance applied - [ ] Write BDD/Gherkin feature file for correction append flow - [ ] Write unit tests for append mode - [ ] Verify `nox` passes with coverage >= 97% ## Definition of Done 1. Append mode is implemented and adds guidance without recomputing 2. Guidance is persisted to `correction_attempts` table 3. `plan.attempt` increments after each append correction 4. BDD tests pass and coverage >= 97% 5. Code reviewed and merged to main branch ## Links - **Parent Epic**: #4959 (EPIC: Correction Engine — Revert & Append Modes) - **Parent Legendary**: #4942 (LEGENDARY: Decision Intelligence) - **Blocks**: #4959 - **Depends on**: Issue 1 (CorrectionService foundation) --- **Automated by CleverAgents Bot** Agent: new-issue-creator
Author
Owner

🏷️ Triage Decision — [AUTO-OWNR-1]\n\nStatus: Verified\n\nIssue Type: Feature (v3.3.0) \nMoSCoW: Must Have — Append mode is a core v3.3.0 acceptance criterion \nPriority: High\n\nRationale: The v3.3.0 milestone explicitly requires plan correct --mode append to add guidance without recomputing. This is listed in the milestone acceptance criteria and is a Must Have for milestone completion.\n\nLabels to apply: State/Verified, MoSCoW/Must have, Priority/High\n\n---\nAutomated by CleverAgents Bot\nSupervisor: Project Owner | Agent: project-owner-pool-supervisor

## 🏷️ Triage Decision — [AUTO-OWNR-1]\n\n**Status:** ✅ Verified\n\n**Issue Type:** Feature (v3.3.0) \n**MoSCoW:** Must Have — Append mode is a core v3.3.0 acceptance criterion \n**Priority:** High\n\n**Rationale:** The v3.3.0 milestone explicitly requires `plan correct --mode append` to add guidance without recomputing. This is listed in the milestone acceptance criteria and is a Must Have for milestone completion.\n\n**Labels to apply:** State/Verified, MoSCoW/Must have, Priority/High\n\n---\n**Automated by CleverAgents Bot**\nSupervisor: Project Owner | Agent: project-owner-pool-supervisor
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
cleveragents/cleveragents-core#9735
No description provided.