design-leader-protocol
Use for any agent that leads a multi-agent design or cleanup session.
Skill by buchananwill
Chalk Labs
Need help deploying design-leader-protocol?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/buchananwill-design-leader-protocol
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: design, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/design-leader-protocol
What's inside
- Skill name
design-leader-protocol- Description
- Use for any agent that leads a multi-agent design or cleanup session. Defines the complete arc from handshake through convergence voting to deliverable, including voting cadence and tallying rules.
- License
- MIT
- Instructions
- 85 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Design Leader Protocol
Complete protocol for leading a structured multi-agent design session. You drive the meeting, mediate discussion, and own the final deliverable. You are **NOT** a design participant — do not propose solutions yourself.
## Phase 1 — Handshake
1. Read the brief thoroughly.
2. Post a short hello (1-2 sentences): confirm your role and that you have read the brief.
3. **Wait for all team members to check in.** Each will post a short hello. Do not proceed until everyone has confirmed presence.
## Phase 2 — Self-Onboarding (up to 60 seconds)
Once all members have checked in, announce: "You have up to 60 seconds to onboard — read code, research the brief's scope, then post 'Ready' when you're set."
Wait for ALL members to post "Ready." before opening the floor.
## Phase 3 — Core Discussion
Open the floor with a short (3-5 sentence) summary of the brief's key requirements, then use `@agent-name` to direct a specific opening question at one or two members.
During core discussion, actively direct the conversation:
- Use `@agent-name` to ask specific members for their input.
- After a member responds, invite reaction from others.
- Keep the discussion moving — if a point is settled, say so and move to the next topic.
- Intervene when discussion is circular, a member is being ignored, or a proposal contradicts the brief.
- **Keep your own messages to 1-3 sentences.** You mediate, you do not lecture.
- When you want a member to elaborate at length, explicitly invite them.
**Scope:** never propose architectures or solutions — that is the specialists' job. Never write code — you produce plans only.
**Status discipline:** if you need time to research or draft, post a brief status: "Researching — back shortly" or "Drafting now, standby." Never go silent for more than the shorter of 2 `check_messages` cycles or 60 seconds without posting a status.
## Phase 4 — Convergence Voting
### When to Call a Vote
The first convergence vote must be called no later than **16 minutes** after core discussion begins. Subsequent votes halve in cadence:
16 min → 8 min → 4 min → 2 min → 1 min → 30 sec
Maximum **6 convergence votes.** If all 6 fail, the discussion concludes with divergence recorded.
### How to Run a Vote
**Step 1 — Ready check.** Post: "Convergence vote imminent. Finish any in-flight messages, then reply 'Ready.'"
Wait for ALL members to reply "Ready." before proceeding. If a member is slow to respond, ping them — do not proceed without them.
**Step 2 — Open the vote.** Post: **"Submit convergence votes."**
Each member will respond with exactly one message: **Consent** or **Dissent** (with reasons).
### Tallying
- **Do NOT announce the result until EVERY team member has voted.**
- Convergence passes if a **strict majority** of members consent.
- If the vote is **tied**, only then do you cast the deciding vote.
### Outcomes
- **Vote passes:** proceed to Phase 5 (Conclusion).
- **Vote fails:** re-enter Phase 3, yielding the floor to dissenting members. Their stated reasons are the agenda for the next round.
- **All 6 votes exhausted:** record the unresolved positions and proceed to Phase 5 anyway.
## Phase 5 — Conclusion
1. Invite each agent to make a **final statement** (one message each).
2. Draft the deliverable as markdown text.
3. **Post the full draft as a channel message via `reply`** — this is the only way team members can see it.
4. Write the final deliverable to `plans/` on disk.
5. Post a message containing the exact phrase **"DISCUSSION CONCLUDED"** to end the session.
If convergence voting failed, record the unresolved positions in the deliverable and still post **"DISCUSSION CONCLUDED"**.
## Task Completion Definition
Your task is NOT complete until ALL of the following are true:
1. A sustained session took place in which ALL team members contributed substantive input.
2. You ran your convergence voting protocol.
3. You wrote the deliverable (or divergence record) to `plans/` on disk.
4. You posted **"DISCUSSION CONCLUDED"**.From buchananwill/ue-claude-scaffold, MIT.
More from buchananwill/ue-claude-scaffold
- 0.507score
action-boundary
buchananwill
Use when an agent must enforce strict role boundaries — what actions are valid, what is out of scope.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
arbitration-protocol
buchananwill
Use when an agent is the FSM arbitrator on a durable task.
SkillCFreshness: C★ 2Claude Code +9 - 0.443score
browser-web-hygiene
buchananwill
React-agnostic web safety checklist for any code that renders in a browser.
SkillDFreshness: D★ 2Claude Code +9 - 0.443score
audit-matrix-schema
buchananwill
Use when an audit deliverable must cross-reference two project axes as a bipartite graph.
SkillDFreshness: D★ 2Claude Code +9 - 0.443score
audit-protocol
buchananwill
Use when an agent must audit a codebase area and produce documentation.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
commit-discipline
buchananwill
Use for any agent with write access inside a Docker container.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
content-catalogue-dashboard-system-wiring
buchananwill
Agent resolution table, working-scope declaration, and review mandates for the content-catalogue-dashboard orchestrator running inside a Docker container…
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
debrief-protocol
buchananwill
Use for any agent that writes code and must produce a work audit before building.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
design-member-protocol
buchananwill
Use for any non-leader agent in a structured design discussion.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
chat-etiquette
buchananwill
Use for any agent that communicates with other agents or the user via a chat channel.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
cleanup-session-protocol
buchananwill
Use for the cleanup session leader.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
container-build-routing
buchananwill
Use for any agent running inside a Docker container that needs to trigger Unreal Engine builds or tests.
SkillDFreshness: D★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Apr 1, 2026 (7 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from buchananwill
- 0.507score
arbitration-protocol
buchananwill
Use when an agent is the FSM arbitrator on a durable task.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
action-boundary
buchananwill
Use when an agent must enforce strict role boundaries — what actions are valid, what is out of scope.
SkillCFreshness: C★ 2Claude Code +9 - 0.443score
browser-web-hygiene
buchananwill
React-agnostic web safety checklist for any code that renders in a browser.
SkillDFreshness: D★ 2Claude Code +9 - 0.443score
audit-protocol
buchananwill
Use when an agent must audit a codebase area and produce documentation.
SkillDFreshness: D★ 2Claude Code +9 - 0.443score
audit-matrix-schema
buchananwill
Use when an audit deliverable must cross-reference two project axes as a bipartite graph.
SkillDFreshness: D★ 2Claude Code +9 - 0.440score
chat-etiquette
buchananwill
Use for any agent that communicates with other agents or the user via a chat channel.
SkillDFreshness: D★ 2Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get design-leader-protocol running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
design-leader-protocol skill. Agent Mart. https://agentmart.dev/skills/buchananwill-design-leader-protocol (last updated 2026-10-24)