Multi-Persona Planning

VerifiedSafe

Collaborative planning method using multiple expert personas to produce a detailed master plan.

Sby Skills Guide Bot
DevelopmentAdvanced
7/22/2026
Claude Code
#planning#multi-agent#architecture#project-scoping#orchestration

Recommended for

Automated summary

This skill runs a multi-persona planning methodology to produce a master plan for a new project or feature.

Strengths

  • Runs a structured multi-phase workflow (scoping, independent plans, debate, revision, synthesis)
  • Generates plans in parallel through multiple expert personas tailored to the project
  • Produces persistent artifacts: per-round plans, debate summary, master plan, and BUILD_STATUS.yaml
  • Offers practical options like --personas, --skip-debate, and --output-dir

Limitations

  • Depends on an external shared/PLANNING_METHODOLOGY.md file for the full reference
  • The workflow includes many approval and wait points, which can slow it down
  • Quality hinges heavily on persona selection and brief clarity
When to use it

Use when starting a new project or a substantial feature and you want a robust master plan drawing on multiple perspectives.

When not to use it

Avoid for small or urgent tasks where multi-round planning would be overkill or too slow.

Security analysis

Safe
Quality score84/100

The skill is a pure orchestration/planning workflow: it parses a brief, reads a user-provided file, writes markdown plans and a BUILD_STATUS.yaml, and spawns sub-agents via the Task tool. There are no destructive commands, no curl|sh, no base64 payloads, no secret/token exfiltration, and no attempts to disable safety controls. File writes are confined to project plan directories. Execution risk is minimal, so it is safe to list.

Findings
  • •Reads a file path specified by the user ('read [file]') as the project brief — benign but unbounded read of any path the user names.
  • •Delegates work to parallel Task sub-agents; those agents inherit the skill's file-read/write instructions but no network, shell, or secret-handling steps are specified.
How this analysis is made

Examples

Thumbnail template
/architect Build a centered-person thumbnail template
Upload pipeline feature
/architect "Add YouTube upload to the podcast pipeline"
Plan from a doc
/architect read evals-course/BUILD_PLAN.md

Skill: /architect

Run the multi-persona planning methodology to produce a master plan for a new project or feature.

Parameters

  • brief (required): What are we building? Can be a sentence, a paragraph, or "read [file]" to pull from an existing doc.
  • --personas (optional): Override persona count. Default: 5.
  • --skip-debate (optional): Skip Phase 2 debate and go straight to synthesis. Faster but lower quality.
  • --output-dir (optional): Where to write plans. Default: auto-detect from project context.

Trigger Phrases

  • /architect Build a centered-person thumbnail template
  • /architect "Add YouTube upload to the podcast pipeline"
  • /architect read evals-course/BUILD_PLAN.md
  • architect a new email drip sequence for cohort 4

Methodology

This skill implements shared/PLANNING_METHODOLOGY.md. The full workflow:

Phase 0  Scope & Persona Selection     → define brief, pick 3-5 expert personas
Phase 1  Independent Plans (Round 1)   → all personas plan in parallel
Phase 2  Debate & Critique             → single moderator resolves conflicts
Phase 3  Revised Plans (Round 2)       → personas revise in parallel
Phase 4  Alignment & Synthesis         → single architect produces master plan
Phase 5  Build Status Tracker          → CREATE BUILD_STATUS.yaml

Execution

1. Parse the brief

If the user provided a file path or "read [file]", read that file as the project brief. Otherwise, use the text they provided directly.

If the brief is too vague (under 20 words, no clear deliverable), ask one clarifying question before proceeding.

2. Determine output directory

Look for context clues:

  • If the brief mentions a specific project (podcast, analytics, evals, etc.), use that project's directory
  • If a working/plans/ directory already exists nearby, use it
  • Otherwise, create working/plans/ in the most relevant project directory
  • If truly ambiguous, ask the user

Set:

  • PLANS_DIR: {project}/working/plans/
  • MASTER_PLAN_PATH: {project}/MASTER_PLAN.md (or {PROJECT_NAME}_MASTER_PLAN.md)

3. Phase 0: Scope & Persona Selection

Read shared/PLANNING_METHODOLOGY.md for the full methodology reference.

Based on the brief, select 3-5 personas. Use the archetype table from the methodology as a starting point, but customize roles to the specific project. For example:

  • A thumbnail project might need: CTR Optimizer, Frontend Renderer, Brand Compositor, Pipeline Architect
  • A course project might need: Curriculum Designer, Student Advocate, Technical Author, Platform Specialist

Present the personas to the user:

Project: [brief summary]
Output: {MASTER_PLAN_PATH}

Personas:
1. [Name] — [Role]. Cares about: [focus]. Will challenge: [what].
2. ...

Proceed with these personas? (a) Yes (b) Swap one out (c) Add/remove

Wait for approval before launching Phase 1.

4. Phase 1: Independent Plans (Round 1)

Launch all persona agents in parallel using the Task tool. Each persona gets:

  • The project brief
  • Their role description and perspective
  • Any reference files or examples mentioned in the brief
  • Instructions to write their plan to {PLANS_DIR}/round1/{persona-slug}.md

Each persona produces:

  1. What needs to be built (their domain)
  2. How it should be structured
  3. Phases/waves
  4. Dependencies on other domains
  5. Risks and unknowns
  6. What they'd push back on

Wait for all personas to complete.

5. Phase 2: Debate & Critique

If --skip-debate: skip to Phase 4.

Launch a single debate agent that receives all Round 1 plans. It identifies:

  • Agreements (2+ personas align)
  • Conflicts (incompatible approaches)
  • Gaps (nobody addressed)
  • Resolutions with reasoning

Output: {PLANS_DIR}/debate-summary.md

6. Phase 3: Revised Plans (Round 2)

Launch all persona agents again in parallel. Each receives:

  • Their Round 1 plan
  • The debate summary
  • Instructions to revise and write to {PLANS_DIR}/round2/{persona-slug}.md

7. Phase 4: Synthesis

Launch a single synthesis agent that receives all Round 2 plans + debate summary.

Produces the master plan with sections:

  1. Executive Summary
  2. Wave Structure (summary table)
  3. Detailed Waves (task specs with IDs, files, deps)
  4. Dependency Graph
  5. Files Changed Summary
  6. Open Questions

Output: {MASTER_PLAN_PATH}

8. Phase 5: Build Status Tracker

After user approves the master plan, generate BUILD_STATUS.yaml following the schema in shared/PLANNING_METHODOLOGY.md.

9. Report

=== PLANNING COMPLETE ===

Master Plan:    {MASTER_PLAN_PATH}
Build Tracker:  {project}/BUILD_STATUS.yaml
Persona Plans:  {PLANS_DIR}/round1/ (5 files)
Revised Plans:  {PLANS_DIR}/round2/ (5 files)
Debate Summary: {PLANS_DIR}/debate-summary.md

Waves: [N]
Tasks: [N]
Ready to execute: "produce wave 0" or read the master plan first

Shortcuts

  • /architect --quick [brief]: Use 3 personas, skip debate (Phases 0-1-4 only). Faster for smaller projects.
  • /architect --resume: Re-read existing plans in working/plans/ and pick up where we left off.

From the file to a first result

The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.

View the source and supporting files

This download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.

1. Download the file

Download SKILL.md

From the root of a test project, save the file as .skill-downloads/multi-persona-planning-mrwak93f/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.

Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.

mkdir -p '.skill-downloads/multi-persona-planning-mrwak93f' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/multi-persona-planning-mrwak93f/download' --output '.skill-downloads/multi-persona-planning-mrwak93f/SKILL.md.part' && mv '.skill-downloads/multi-persona-planning-mrwak93f/SKILL.md.part' '.skill-downloads/multi-persona-planning-mrwak93f/SKILL.md'

2. Check prerequisites

Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.

No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.

The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.

3. Load the skill

This file does not provide a usable name/description header for this native procedure. It remains available to download and inspect; use the manual method below.

Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.

4. Try a first task

Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:

Read .skill-downloads/multi-persona-planning-mrwak93f/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.

Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.

Related skills