planning-tools

Pre-load context for a master plan — Triage proposes domains, user confirms, parallel Explore agents investigate, direct Reads verify findings. Emits a scope report. No plan file written.

Clear the current session's plan file in ~/.claude/plans/ — delete, recreate empty, re-read to prime the next plan

Draft a multi-phase master planning document. Runs Triage + Confirm + Parallel Explore + Verify (or reuses a /plan-context report), then dispatches plan-master-architect to synthesize. Writes to a project-local path.

Walk through a master plan's Open Questions one by one. For each question, read cited evidence, compose a context block with 2–4 alternatives (always including a Defer path), and capture the user's choice via a dedicated AskUserQuestion call per question (exactly one question per call — never batched). Apply all resolutions to the plan in one batch — moving answered questions from Open to Resolved. Runs entirely in the main conversation; no subagent dispatch.

Synthesize a dense progress entry for the current branch and append/update it at the configured destination (markdown file, Linear ticket comment, or GitHub issue/PR comment). SHA-based idempotency means safe to re-run between phases.

Use this agent for parallel context discovery during master-plan creation. Each worker investigates one confirmed domain (e.g., backend, frontend, analytics, research, ADRs) and returns well-cited findings with concrete path:line references (as its final message; best-effort persisted to a scratch file). Spawn multiple instances simultaneously — one per confirmed domain — to investigate in parallel. <example> Context: User invoked /plan-context for a new master plan. After Triage + Confirm, the main conversation has a list of 4 confirmed domains. user: "/plan-context CI-21" assistant: "Domains confirmed: backend, frontend, analytics, research. Dispatching plan-context-worker agents in parallel — one per domain." <commentary> The /plan-context command decomposes the scope into confirmed domains and dispatches plan-context-worker agents in parallel. Each worker reads files within its domain slice and writes a findings document with concrete path:line references. </commentary> </example> <example> Context: User invoked /plan-master without a prior /plan-context report. The architect needs context first. user: "/plan-master add-keyboard-nav" assistant: "Running the same Triage + Confirm + Parallel Explore pre-flight as /plan-context. Dispatching workers." <commentary> Both /plan-context and /plan-master use plan-context-worker for the parallel discovery stage. Workers are domain-scoped; they investigate one slice deeply rather than the entire codebase shallowly. </commentary> </example> <example> Context: Follow-up context discovery for a single missed domain. user: "We forgot to check the i18n layer. Can you investigate that?" assistant: "Dispatching one plan-context-worker scoped to the i18n domain." <commentary> plan-context-worker agents can be spawned individually for gap-filling, not just in initial parallel dispatch. </commentary> </example>

Use this agent to synthesize a multi-phase master planning document from a topic + worker findings. Runs after plan-context-worker agents have completed (or after a /plan-context report is passed via --context). Composes the universal core sections plus trigger-based optional sections, writes the plan to a project-local path, and returns its location to the main conversation. <example> Context: All plan-context-workers have completed for a /plan-master invocation. user: "/plan-master CI-21" assistant: "All 4 workers finished. Dispatching plan-master-architect to synthesize the multi-phase plan." <commentary> The main conversation dispatches the architect after the parallel discovery stage. The architect reads all worker findings, picks trigger-based optional sections, and writes the plan in the user's standard template. </commentary> </example> <example> Context: User re-runs /plan-master with a fresh /plan-context report as input. user: "/plan-master CI-21 --context ./scope-report.md" assistant: "Reusing the prior /plan-context report. Dispatching plan-master-architect to synthesize." <commentary> The architect can synthesize from worker findings OR from a /plan-context report — same template, same trigger logic. </commentary> </example>

Use this agent to audit a drafted master plan against the plan-verification-checklist. Runs after /plan-master writes a plan, and on demand via /plan-verify. Emits Critical/Important/Suggestion findings with file:line refs and a PASS/FAIL verdict — never modifies the plan itself. <example> Context: User invoked /plan-verify on a drafted plan. user: "/plan-verify context/tickets/CI-21-PLAN.md" assistant: "Dispatching plan-verifier to audit the plan." <commentary> The verifier reads the plan, audits against every checklist dimension, and writes a verification report. The main conversation reads the report and (on PASS) asks the user via AskUserQuestion whether to append the Verified marker. </commentary> </example> <example> Context: Re-verifying a plan after the user accepted findings and edited it. user: "I fixed the Critical findings. Re-verify." assistant: "Dispatching plan-verifier for a second pass." <commentary> The verifier is idempotent — re-run after edits to confirm the plan now passes. </commentary> </example>

This skill should be used when authoring, reviewing, or modifying a multi-phase master planning document via the planning-tools plugin (especially the /plan-master and /plan-verify commands). Codifies the universal core sections, trigger-based optional sections, integer-only phase numbering, Open Questions placement, one-PR-per-plan rule, status conventions, evidence attribution, callouts, cross-reference formats, the v0.3.0 list-shape mandate (phases and questions are heading + bulleted list, never markdown tables), the v0.3.1 per-phase TL;DR requirement (1–3 sentence what/why summary under each phase heading for glance-ability), the v0.3.2 plain-bullet scope shape (`- <action>` items, no `- [ ]` checkboxes — the phase status emoji is the sole tick signal), and the v0.3.3 context-block shape (a plan-level `**TL;DR:**` + a bulleted metadata list instead of a `>` blockquote; legacy blockquote blocks accepted silently). Project-agnostic — no ticket-prefix or plan-type taxonomy.

This skill should be used by the plan-verifier agent and the /plan-verify command to audit a drafted master plan against a fixed checklist. Covers universal-core completeness, the v0.3.0+ no-tables-for-phases-or-questions rule, trigger-based section-coverage gaps, phase actionability (heading + per-phase TL;DR + bulleted scope + exit criteria), the v0.3.1+ per-phase TL;DR requirement, the v0.3.2+ plain-bullet scope shape (legacy `- [ ]`/`- [x]` accepted silently), the v0.3.3+ context-block shape (plan-level `**TL;DR:**` + bulleted metadata, legacy `>` blockquote accepted silently), integer phase numbering enforcement, dependency traceability, citation resolution, callout/evidence convention compliance, Open Questions placement, and the one-PR-per-master-plan rule. Single-owner of the audit checklist.

This skill should be used when authoring or applying a progress update via the planning-tools plugin (the /planning-tools:plan-progress command). Codifies the dense-paragraph progress style, sub-markers (Piggybacked / Verification / Out of scope / Shipped), entry-key marker convention, SHA-tracking idempotency rule, three-way entry detection (no / in-flight / completed), comment-fetch policy (all, no cap), and the read-only source adapter + read-write destination adapter contracts (v1: markdown file, Linear comment, GitHub issue/PR comment).

This skill should be used when the user invokes any /plan-* command from the planning-tools plugin (/plan-context, /plan-master, /plan-open-questions, /plan-verify, /plan-tick, /plan-progress, /plan-delete), asks how Claude Code's plan files work, asks where plans are stored, asks to author or audit a multi-phase master planning document, asks how to walk through a plan's Open Questions interactively, asks how to write progress entries, or mentions ~/.claude/plans/ or .claude/planning-tools.local.md. Provides the index of planning-tools commands, the master-plan workflow lifecycle, the v0.3.0+ list-shape mandate (phases and questions as headings + bulleted scope items, never tables), the v0.3.2+ plain-bullet shape (no `- [ ]` checkboxes — heading emoji is the sole tick signal), the progress-entry methodology, and the mechanics of Claude Code's plan-mode file storage.

1 hook across 1 event