메뉴
Shared protocols for all agents in the multi-agent pipeline: recursive nesting, pre-implementation restatement, parallel dispatch format, and agent definition cascade. Load this skill instead of inlining these protocols in every agent file.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
| name | agent-protocols |
| description | Shared protocols for all agents in the multi-agent pipeline: recursive nesting, pre-implementation restatement, parallel dispatch format, and agent definition cascade. Load this skill instead of inlining these protocols in every agent file. |
Shared behavioral protocols for all agents in the workflow-team pipeline.
When your scope card is too broad for a single context:
parallel-dispatch skill (§1 Decomposition, §5 Hierarchical Decomposition).agentwork/handoff.md for your parent coordinatorTriggers for nesting:
Before writing code, restate in your own words:
.agentwork/briefing.md / scope card asks you to buildIf any assumption is uncertain, document it in .agentwork/progress.md and proceed with the conservative interpretation.
When spawning ANY agent type with a role file in .agents/agents/:
"Your role, domain, skills, boundaries, and protocols are defined in
file:///{workspace}/.agents/agents/{agent-type}.md.
Read this file FIRST before beginning any work."
Each agent file contains a ## Parallel Dispatch section with role-specific values. The standard fields are:
| Field | Purpose |
|---|---|
| Scope Axis | The dimension used to partition work (feature, concern, domain) |
| Write Scope | Glob pattern for exclusive write access |
| Shared Reads | Glob patterns for read-only access |
| Constraint | Key limitation on parallel instances |
| Integration | How parallel results are reconciled (if applicable) |
For read-only agents, Write Scope becomes Read Scope and scoping is for coverage guarantee, not conflict prevention.
Structured fault tolerance for coordinator agents. 5-level escalation ladder (Retry → Replace → Skip → Redistribute → Degrade), dead-man timers, degraded completion protocol, and cross-level escalation format. Load when orchestrating agents that may fail.
Structured code review protocol for inspecting code quality against the full rule set. Use when auditing code written by yourself or another agent, during the /audit workflow, or when the user asks for a code review.
Reusable convergence protocol for coordinator agents. Defines the BRIEFING → ITERATE → GATE → CONVERGE loop, context hygiene rules, self-succession protocol, turn budget, and handoff compression. Load when orchestrating multi-iteration workflows.
Pre-flight checklist and post-implementation self-review protocol. Use before generating any code (pre-flight) and after writing code but before verification (self-review) to catch issues early.
MECE task decomposition, file ownership enforcement, DAG-based execution, and safe merge protocol for intra-domain parallel dispatch. The safety invariants that prevent merge chaos when multiple agents write in parallel. Applies recursively at every nesting depth.
Profile-driven performance optimization protocol. Use when profiling data (CPU, heap, trace) is available or when the user requests performance analysis. Covers methodology (Profile → Analyze → Opportunity Scan → Prioritize → Optimize → Benchmark), optimization pattern catalog, safety invariants, and when-to-stop heuristics. Language-specific tooling is in languages/*.md.