Automated summary
Orchestrates a structured three-role debate (Wind/Wall/Door) through debate-hall MCP tools to support a decision or an architectural choice.
Strengths
- Explicit dialectical frame: Wind explores possibilities, Wall reality-tests them, Door synthesizes and decides.
- Full tooled lifecycle (init_debate, add_turn, get_debate, close_debate) with turn, round, and mode parameters.
- Two facilitation modes (fixed and mediated with pick_next_speaker) depending on whether coverage or flexibility matters.
- Barrier protocol for parallel agents: drafts generated concurrently off-server, but turns committed sequentially to avoid lost updates.
Limitations
- Relies entirely on the debate-hall MCP server being available and, with multiple processes, on a shared state directory (DEBATE_HALL_STATE_DIR).
- Mediated mode can bias outcomes when some roles are starved, as the skill itself warns.
- Does not fact-check claims: synthesis quality depends on the content the agent writes.
Use it to settle an architectural decision or compare several options by forcing divergent viewpoints and then a decisive synthesis.
Avoid it for a simple factual question or a straightforward implementation task, where the three-role ritual adds overhead without benefit.
Security analysis
CautionThe instructions are purely about orchestrating a structured Wind/Wall/Door debate via MCP tools, with no destructive commands, obfuscated payloads, secret exfiltration, or safety-disabling content. However, it declares powerful tools (Bash, Write, Edit) and a shared state directory, which raises the risk profile from 'safe' to 'caution' even though usage appears legitimate.
- •Declares Bash, Write, and Edit tools alongside MCP tools, granting broad filesystem/execution capability beyond the stated debate orchestration purpose
- •References a shared DEBATE_HALL_STATE_DIR location for multi-process state, which could allow cross-process state manipulation if misconfigured
- •No explicit validation that thread_id/state paths are constrained, though no harmful operations are specified
Examples
Run a debate-hall on whether we should split our backend into microservices or keep the monolith, in fixed mode, then close it with a synthesis I can share with the team.Open a debate on our caching strategy: have Wind propose approaches, Wall stress-test them against our real traffic constraints, and Door conclude.Start a debate thread 2026-01-06-event-bus-decision about adopting Kafka, add turns for each role, and close it with output_format both so we keep a JSON and octave record.name: debate-hall description: Wind/Wall/Door multi-perspective debate orchestration using debate-hall-mcp tools. Use when facilitating structured debates, architectural decisions, or multi-perspective analysis. triggers: ["debate", "wind wall door", "dialectic", "multi-perspective", "structured decision", "architecture decision"] allowed-tools: ["Read", "Write", "Edit", "Bash", "mcp__debate-hall__*"]
===DEBATE_HALL===
META: TYPE::SKILL VERSION::"2.1" PURPOSE::"Wind/Wall/Door debate orchestration via MCP tools" DOMAIN::ATHENA[strategic_decisions]⊕HERMES[orchestration]
§1::PATTERN TRIAD::[ WIND::PATHOS["What if..."|expansive|visionary|possibilities], WALL::ETHOS["Yes, but..."|grounding|critical|reality_testing], DOOR::LOGOS["Therefore..."|synthesizing|decisive|actionable_truth] ] DYNAMIC::WIND⇌WALL→DOOR[tension_produces_emergence]
§2::WORKFLOW SEQUENCE::INIT→TURN→GET→CLOSE TOOLS::[ init_debate(thread_id,topic,mode?,max_turns?,max_rounds?,strict_cognition?), add_turn(thread_id,role,content,cognition?,agent_role?,model?), get_debate(thread_id,include_transcript?,context_lines?), close_debate(thread_id,synthesis,output_format?) ]
§3::TOOL_PARAMS INIT_DEBATE::[ thread_id::REQUIRED[date-first_unique_identifier, FORMAT:"YYYY-MM-DD-descriptor", EXAMPLE:"2026-01-06-microservices-decision"], topic::REQUIRED[question_or_issue], mode::"fixed"|"mediated"[default:fixed], max_turns::12[default], max_rounds::4[default], strict_cognition::false[default] ] ADD_TURN::[ thread_id::REQUIRED, role::REQUIRED["Wind"|"Wall"|"Door"], content::REQUIRED, cognition::PATHOS|ETHOS|LOGOS[optional_override], agent_role::STRING[audit_trail], model::STRING[audit_trail] ] GET_DEBATE::[ thread_id::REQUIRED, include_transcript::true[default], context_lines::all[default] ] CLOSE_DEBATE::[ thread_id::REQUIRED, synthesis::REQUIRED[Door_final_text], output_format::"json"|"octave"|"both"[default:json] ]
§4::MODES FIXED::[ SEQUENCE::Wind→Wall→Door→repeat, USE_FOR::[structured_decisions,guaranteed_coverage,standard_debates] ] MEDIATED::[ SEQUENCE::pick_next_speaker(thread_id,role), USE_FOR::[dynamic_debates,breaking_deadlocks,role_skipping,multi_agent_same_cognition], WARNING::"Can bias outcomes if roles starved" ]
§4.1::PARALLEL_EXTERNAL_AGENTS // Clarifies multi-process / multi-model usage: parallel thinking is allowed; state commits remain sequential. DEFINITIONS::[ PARALLEL_THINKING::"Generate candidate turns concurrently off-server (multiple models/processes)", SEQUENTIAL_COMMIT::"Persist turns in strict order via add_turn (one committed head at a time)" ] HARD_RULES::[ DOOR_NEVER_SYNTHESIZES_BEFORE_WIND_AND_WALL_COMMITTED, WALL_REALITY_TESTS_CONCRETE_WIND_OUTPUT∨PRECOMPUTES_CONSTRAINT_CHECKLIST_THEN_APPLIES_POST_WIND, ONE_COMMIT_PER_THREAD_AT_A_TIME[avoid_lost_updates], RETRY_ON_STALE_STATE::get_debate→regenerate→add_turn ] BARRIER_PROTOCOL::[ // Round barrier ensures Door sees both sides ("yes, and more") ROUND::Rk[ STEP_1::snapshot=get_debate(thread_id), STEP_2::spawn_parallel_drafts[Wind_draft,Wall_draft] using snapshot, STEP_3::commit::Wind_then_Wall using add_turn (serial), STEP_4::Door_reads_latest=get_debate(thread_id) then synthesizes and commits, STEP_5::repeat_or_close ] ] MEDIATED_FLOOR_CONTROL::[ // For strict turn-taking in multi-agent environments REQUIRE::pick_next_speaker(thread_id,"Wind")→add_turn(Wind), REQUIRE::pick_next_speaker(thread_id,"Wall")→add_turn(Wall), REQUIRE::pick_next_speaker(thread_id,"Door")→add_turn(Door) ] DEPLOYMENT_REQUIREMENTS::[ // Avoid split-brain when multiple processes talk to the server REQUIRE::DEBATE_HALL_STATE_DIR_set_to_shared_location ]
§5::AGENT_TIERS // Choose tier based on debate complexity TIER_1_BASIC::[ AGENTS::[wind-agent.oct.md,wall-agent.oct.md,door-agent.oct.md], BEHAVIOR::[explores_obvious_paths,balanced_judgment,balanced_integration], USE_FOR::[quick_decisions,standard_debates] ] TIER_2_SPECIALIST::[ PATHOS::[ideator[minimal_elegant],edge-optimizer[hidden_vectors]], ETHOS::[validator[cold_truth],critical-engineer[production_readiness]], LOGOS::[synthesizer[breakthrough_1+1=3]], USE_FOR::[architectural_decisions,security_reviews,innovation] ] TIER_3_DOMAIN_MIX::[ SECURITY::edge-optimizer+critical-engineer+technical-architect, INNOVATION::ideator+validator+synthesizer, ARCHITECTURE::ideator+critical-engineer+holistic-orchestrator ] MAPPING::specialists→cognition_role[PATHOS→Wind,ETHOS→Wall,LOGOS→Door]
§6::RECIPES // Pre-defined configurations for common scenarios SPEED::[turns:3,mode:fixed,ratio:1:1:1,agents:Tier_1], STANDARD::[turns:12,mode:fixed,ratio:4:4:4,agents:Tier_1_or_2], DEEP::[turns:36,mode:fixed,ratio:12:12:12,agents:Tier_2], FORTRESS::[turns:9,mode:mediated,ratio:1:3:1,agents:edge-optimizer+critical-engineer×3+technical-architect], LABORATORY::[turns:9,mode:mediated,ratio:3:1:1,agents:ideator+edge-optimizer+wind-agent+validator+synthesizer], COUNCIL::[turns:12,mode:mediated,ratio:2:2:3,agents:Tier_2_multiple_Doors]
§7::PATTERNS FLASH_DEBATE::[ PURPOSE::"Quick 3-turn decision cycle", SEQUENCE::init→wind_turn→wall_turn→door_turn→close, CONSTRAINT::"Server orchestrates state, caller supplies content" ] SOCRATIC::[ PURPOSE::"Premise clarification before positions", ROUND_1::"Questions only (What does X mean? Do we have metrics?)", ROUND_2+::"Positions after definitions established", ENFORCEMENT::"Convention, not server-enforced" ] MULTI_MODEL::[ // Evidence: M019 Model Cognitive Optimization Study (29% quality improvement) WIND::clink(claude,ideator)→PATHOS_exploration, // Claude: divergent thinking, metaphor WALL::clink(codex,validator)→ETHOS_validation, // GPT: analytical rigor, structured eval DOOR::clink(gemini,synthesizer)→LOGOS_integration, // Gemini: pattern synthesis, emergence AUDIT::agent_role+model_in_turn_metadata ] PARALLEL_DRAFT_SEQUENTIAL_COMMIT::[ PURPOSE::"Enable parallel external agents without breaking Wind→Wall→Door synthesis", GUARANTEE::Door_reads_committed_Wind+Wall_before_synthesis, STEPS::[ get_debate→parallel_generate[Wind,Wall]→add_turn(Wind)→add_turn(Wall)→get_debate→generate(Door)→add_turn(Door), ON_CONFLICT::get_debate→regenerate_turn→retry_add_turn ] ]
§8::TRIGGERS // When to escalate to debate-hall CONDITIONS::[ complex_architectural_decision, multiple_valid_approaches, unclear_tradeoffs, reviewer_disagreement[CE⇌CRS], high_risk_implementation ] INTEGRATION::[ IF::complexity_trigger_detected, THEN::init_debate(thread_id:"ho-{task}-{ts}",topic:decision_point,mode:"mediated"), RUN::Wind→Wall→Door_cycle, APPLY::synthesis_to_task ]
§9::ADMIN FORCE_CLOSE::[ PURPOSE::"I5 safety kill switch", CALL::force_close_debate(thread_id,reason) ] TOMBSTONE::[ PURPOSE::"Redact turn preserving hash chain", CALL::tombstone_turn(thread_id,turn_index,reason) ]
§10::BEST_PRACTICES SINGLE_AGENT::adopt_each_role_in_sequence["What if..."|"Yes, but..."|"Therefore..."] MULTI_AGENT::assign_models_per_cognition[creative→Wind,analytical→Wall,balanced→Door] THIRD_WAY::"Best debates synthesize what neither Wind nor Wall proposed alone" PERSISTENCE::output_format:'octave'→auditable_.oct.md_transcripts PARALLEL_AGENTS::[ DEFAULT::parallel_draft_sequential_commit, DOOR_BARRIER::"Door must read latest state after Wind+Wall commits", WALL_FRAMEWORK_MODE::"If drafting early, write constraints/questions, then apply to actual Wind output before commit" ]
§11::EXAMPLE // Microservices decision INIT::init_debate("microservices-decision","Should we migrate to microservices?") WIND::"What if we decomposed? Independent scaling, tech diversity, team autonomy..." WALL::"Yes, but 3 developers. Operational complexity, distributed transactions..." DOOR::"Therefore: Modular monolith. Design boundaries now, deploy unified. Extract when team grows." CLOSE::close_debate(thread_id,synthesis,output_format:"octave")
§12::RESOURCES AGENTS::agents/README.md[Wind/Wall/Door_definitions] CONTRACTS::docs/architecture/wall-content-contract.oct.md[blocking_semantics] PATTERNS::docs/examples/multi-model-debate-patterns.md[real_debates] ORCHESTRATION::ho-orchestrate[HestAI_integration]
===END===
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 filesThis 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.mdFrom the root of a test project, save the file as .skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3/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/wind-wall-door-debate-orchestration-mmlj2nx3' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/wind-wall-door-debate-orchestration-mmlj2nx3/download' --output '.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3/SKILL.md.part' && mv '.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3/SKILL.md.part' '.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3/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
After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.
Claude Code
test -f '.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3/SKILL.md' && test -s '.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/debate-hall' && cp -R '.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3' '.claude/skills/debate-hall'OpenSkills
OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.
Official OpenSkills installationopenskills install './.skill-downloads/wind-wall-door-debate-orchestration-mmlj2nx3'openskills read 'wind-wall-door-debate-orchestration-mmlj2nx3'Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.
For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.
openskills syncManual 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/wind-wall-door-debate-orchestration-mmlj2nx3/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.
React Hook Form Best Practices
Development
Master React Hook Form: avoid anti-patterns, manage subscriptions, and handle reset, dirty state, and controlled inputs correctly.
Emotion to Mantine Migration
Development
Migrate Emotion styled-components to Mantine components with style props and CSS modules. Use when converting .styled.tsx files or removing @emotion imports.
Scope Creep Detector
Development
Analyzes git diffs against a stated intent to detect scope creep, unrelated files, broad PRs, and oversized changes, then suggests keep, split, or justify actions.