# Delivery Planning > Use when delivery work needs a proposal before editing task candidates, Kanban cards, backlog shape, or board changes. - **Type:** Skill - **Install:** `agentstack add skill-choral-io-choral-skills-delivery-planning` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [choral-io](https://agentstack.voostack.com/s/choral-io) - **Installs:** 0 - **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [choral-io](https://github.com/choral-io) - **Source:** https://github.com/choral-io/choral-skills/tree/main/skills/delivery-planning - **Website:** https://github.com/choral-io/choral-skills ## Install ```sh agentstack add skill-choral-io-choral-skills-delivery-planning ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Delivery Planning ## Runtime Context Before acting, resolve `` using the runtime bootstrap rules, then read `/.workflow/runtime.md` and `/.workflow/manifest.yml`; do not assume non-default workflow paths or default ids. Use this skill to propose Kanban changes from project knowledge. This skill produces a dry-run only. Task items are candidates and context until an approved Kanban card links to them. Use this skill when the user asks to rank loose task items, review backlog candidates, or turn task items into accepted Kanban work. ## Workflow 1. Read `/.workflow/rules/delivery.md`. 2. Read `/.workflow/schemas/common.md` and `/.workflow/schemas/tasks.md`. 3. Collect candidate task items and source knowledge. 4. Exclude local workspace notes, archived notes, and localized files. 5. When proposing assignees, reviewers, ownership fit, or handoffs, read only the relevant sections from `/members/.md`, such as `Responsibilities`, `Focus Areas`, `Collaboration`, or `Availability`. 6. De-duplicate candidates against `/planning/KANBAN.md`. 7. Produce a dry-run table and wait for maintainer approval. ## Default Inputs - `/product/**` - `/user-stories/**` - `/discovery/**` - `/concepts/**` - `/architecture/**` - `/decisions/**` - `/test-cases/**` - `/metrics/**` - `/experiments/**` - `/releases/**` - `/guidelines/**` - `/tasks/*.md` Use `/guidelines/**` as planning context or constraints. Do not create Kanban candidates from guidelines alone unless a guideline explicitly defines executable delivery work. Use `/discovery/**` as supporting evidence, assumptions, and opportunity context for product requirements. Do not create Kanban candidates from discovery analysis alone unless it has been accepted as product scope or a task item. Use `/user-stories/**` as source intent for user goals, use cases, scenarios, journeys, and acceptance intent. Do not create Kanban candidates from stories alone unless they are linked to accepted product scope or converted into task items. Use `/test-cases/**` as validation context for task readiness and acceptance coverage. Do not create Kanban candidates from test cases alone unless the test case explicitly identifies executable delivery work. Use `/metrics/**`, `/experiments/**`, and `/releases/**` as planning context, evidence, constraints, or follow-up signals. Do not create Kanban candidates directly from raw metric movement, running experiments, or release notes unless a task item or accepted scope exists. Use `/proposals/**` only as backlog review context. Do not create Kanban candidates directly from proposals. Accepted task proposals must be converted into task items before delivery planning. Use `type: issue`, `type: bug`, and `type: defect` task items as delivery candidates only after triage shows they are actionable. Do not plan raw feedback, unverified observations, duplicates, invalid reports, or unresolved non-reproducible defects as Ready work. When decomposing one requirement into several dependent tasks, propose dependent tasks for `Backlog` by default. Use `blocked_by` and `readiness: blocked` for planned dependency order; propose `Ready` only when every Ready Checklist source requirement is satisfied and `owners` resolves to existing member or group documents. Use `/workspace/*/summaries/**`, `/workspace/*/handoffs/**`, or `/workspace/*/research/**` only when the owner or maintainer explicitly selects it. Never use `/workspace/*/local/**` as planning input. Do not read member local workspace instructions for team planning. ## Output - Concise summary of proposed cards. - Dry-run table. - Missing metadata or blockers. ## References - For the dry-run table and task metadata examples, read `references/dry-run.md`. ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [choral-io](https://github.com/choral-io) - **Source:** [choral-io/choral-skills](https://github.com/choral-io/choral-skills) - **License:** MIT - **Homepage:** https://github.com/choral-io/choral-skills Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-choral-io-choral-skills-delivery-planning - Seller: https://agentstack.voostack.com/s/choral-io - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.