beesdk-plan-iteration

Validates BeeSDK tasks against repository evidence and produces roadmap items and copy-ready Issues.

Updated Aug 24, 2026
One-click install
npx skills add https://github.com/beesyst/beesdk --skill beesdk-plan-iteration-beesyst
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: beesdk-plan-iteration
Source: https://github.com/beesyst/beesdk/tree/main/.agents/skills/beesdk-plan-iteration
Command: npx skills add https://github.com/beesyst/beesdk --skill beesdk-plan-iteration-beesyst

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Planning a BeeSDK iteration often relies on stale roadmaps, unverified assumptions, or unclear ownership between BeeSDK and consumer repositories. This Skill inspects the actual implementation through Bee Dev MCP, validates whether work is truly necessary, and produces a reconciled roadmap item plus complete, copy-ready Issues without modifying any repository. ## Core Features & Use Cases - Evidence-based validation: Treats tasks and roadmap references as hypotheses and verifies them against current code, tests, contracts, and package metadata via Bee Dev MCP. - Roadmap reconciliation: Chooses exactly one planning decision (reuse, refine, replace, insert, standalone, or reject) while preserving completed history and numbering rules. - Ownership and cross-repository planning: Determines whether work belongs to BeeSDK, a consumer, a domain module, or BeeUI, and defines implementation and release order per repository. - Use Case: A user proposes "Iteration SDK-4" for a new public contract. The Skill inspects the repo, finds the capability already exists, rejects the work as unnecessary, and explains the simpler alternative with evidence. ## Quick Start Ask the assistant to run the BeeSDK planning workflow for your task or idea, providing the main worktree path, mode, roadmap context, and any additional repositories.

Frequently Asked Questions about beesdk-plan-iteration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I plan a BeeSDK iteration before implementation?

Run this planning workflow with your task or idea, main worktree path, mode, and roadmap context. It inspects the repository through Bee Dev MCP, validates necessity and ownership, and returns a roadmap fragment plus a complete copy-ready Issue.

How does the skill decide between a numbered iteration and a standalone task?

Numbered iterations are required for changes to public contracts, package behavior, typing, compatibility, or cross-repository contracts. Standalone classification applies only to narrow maintenance like typo fixes, small test corrections, or housekeeping without public API impact.

Does this workflow modify repositories or create Issues automatically?

No. The workflow is strictly read-only: it never modifies files, switches branches, runs tests, or creates Issues, branches, commits, or PRs. It only returns copy-ready roadmap and Issue text for manual use.

What happens if the referenced roadmap iteration is already done?

Completed items are never returned to PLANNED or renumbered. The workflow checks whether the request is already implemented and only creates a new unique ID for genuine follow-up work, using decimal IDs solely for direct continuations.

When should work go to a consumer repository instead of BeeSDK?

Runtime implementation, orchestration, module registries, storage, connectors, and product workflows belong to consumer repositories. BeeSDK owns only shared public contracts, protocols, typing, and package build behavior, and the workflow rejects plans that violate these boundaries.

What are the limitations of this planning workflow?

It cannot execute or verify code, depends on Bee Dev MCP access to the exact worktrees, and returns PLANNING INCOMPLETE when mandatory evidence cannot be read. It also never prepares implementation prompts, which require a separate prompt-02 run per repository.