huacheng

Process file annotations — triage, cross-impact assessment, and execution. Triggered automatically when annotations (Insert/Delete/Replace/Comment) are submitted from the file viewer UI via JSONL prompt.

Six-dimension gated review — evaluates plans, implementations, and skills through D1-D6 quality gates with convergence tracking. Use for post-plan review, mid-exec assessment, post-exec acceptance, or any time the user says 'check', 'review', 'audit', 'evaluate quality', or wants to know if deliverables are ready.

Execute the implementation plan — write code, make changes, build the deliverables. Use when the user says 'do it', 'implement', 'start coding', 'execute the plan', 'build it', or wants to begin actual implementation work after planning is complete.

Cross-task knowledge library management — search, list, audit, maintain, and evolve the shared .library/ knowledge base. Includes security rules evolution loop (separate from task auto loop).

Generate step-by-step implementation plans with verification criteria. Use when the user asks 'how should we do this', 'make a plan', 'break this down into steps', 'what's the approach', or needs a roadmap before implementation.

Investigate requirements and domain knowledge to support any lifecycle phase. Default mode: autonomous objective refinement (O1→O2→O3) with background research, feasibility analysis, and goal synthesis — all three stages completed in one pass. Also callable from plan, exec, verify, or check for gap-only reference collection. Use when the user wants to understand before acting — 'research this', 'what are the options', 'feasibility analysis', or 'deepen the objective'.

Define task objectives and requirements in .target.md. Use when the user describes what they want to build, says 'I want to...', 'the goal is...', 'set the target', 'define requirements', or needs to clarify/refine what should be accomplished before planning.

Run domain-adapted tests and verification procedures, producing result files. Triggered after plan generation (post-plan), during execution (per-step), or after execution completes (post-exec) to provide evidence for check's verdict.

Process Plan panel annotations — triage, cross-impact assessment, and execution. Triggered automatically when annotations (Insert/Delete/Replace/Comment) are submitted from the Plan panel UI.

Autonomous execution loop — single Claude session orchestrates plan/check/exec cycle internally

Cancel a task module — set status to cancelled, stop auto if running, optionally clean up worktree. Use when a task becomes infeasible, is deprioritized, or needs to be abandoned.

Check plan feasibility at key checkpoints — post-plan, mid-execution, post-execution

Execute the implementation plan for a reviewed task module. Triggered after check PASS (from review status) or on NEEDS_FIX continuation (from executing status with fix guidance).

Initialize a new task module in AiTasks/ directory with system files, git branch, and optional worktree. Use when starting a new feature, bug fix, or refactoring task that needs structured lifecycle tracking.

Query task status and relationships — read-only, no file writes. Use to inspect task inventory, check progress, view dependency graphs, or audit status transition history.

Merge completed task branch to main — with conflict resolution, verification retry, and cleanup. Triggered after check post-exec ACCEPT verdict confirms all tests pass.