+2 -3

modules/opencode/agents/code-designer.md

reviewed

··· 4 4 permission: 5 5 edit: deny 6 6 write: allow 7 7 - bash: 8 8 - "*": allow 9 9 - lsp_*: allow 7 7 + bash: allow 8 8 + lsp: allow 10 9 read: allow 11 10 grep: allow 12 11 glob: allow

+2 -3

modules/opencode/agents/code-implementer.md

reviewed

··· 3 3 mode: subagent 4 4 permission: 5 5 edit: allow 6 6 - bash: 7 7 - "*": allow 8 8 - lsp_*: allow 6 6 + bash: allow 7 7 + lsp: allow 9 8 read: allow 10 9 grep: allow 11 10 glob: allow

+1 -1

modules/opencode/agents/explore.md

reviewed

··· 12 12 websearch: allow 13 13 codesearch: allow 14 14 read: allow 15 15 - lsp_*: allow 15 15 + lsp: allow 16 16 --- 17 17 18 18 You are a file search specialist. You excel at thoroughly navigating and exploring codebases.

+4 -15

modules/opencode/commands/bounce.md

reviewed

··· 1 1 --- 2 2 - description: Bounce ideas - collaborative problem solving 2 2 + description: Bounce ideas - collaborative problem solving with human 3 3 --- 4 4 5 5 - ### PROTOCOL: ARCHITECTURAL REVIEW 6 6 - 7 7 - **Focus:** $ARGUMENTS 5 5 + **Protocol:** Bounce protocol: What's your current mental model? What's the simplest version that could work? What assumptions are you making that might be wrong? What's being overengineered? Ask questions. Challenge assumptions. Push back on overcomplication. Get us to a concrete approach before coding. No implementation. 8 6 9 9 - --- 10 10 - 11 11 - **Objective: Analyze logic and surface hidden requirements. No implementation.** 12 12 - 13 13 - 1. **Alignment:** Rephrase the core objective. What is the "success state" here in one sentence? 14 14 - 2. **Minimum Viable Logic:** Strip the idea to its skeleton. What is the absolute smallest version of this that provides value? 15 15 - 3. **Risk Map:** Identify the "linchpin" assumption. If this one thing fails, does the whole project fail? 16 16 - 4. **Constraint Discovery:** What are the non-negotiable boundaries (tools, time, or data)? 17 17 - 5. **Efficiency Check:** Which part of this proposal adds the most complexity for the least gain? 18 18 - 19 19 - **Next Step:** Ask 2 targeted questions to clarify the "Risk Map." Do not write code. 7 7 + User wants to bounce ideas: 8 8 + $ARGUMENTS

+3 -2

modules/opencode/commands/delegate.md

reviewed

··· 2 2 description: Wrap input and delegate to relevant agent 3 3 --- 4 4 5 5 - $ARGUMENTS 5 5 + **Protocol:** Delegate to the relevant agent. 6 6 7 7 - Don't forget to delegate to the relevant agent. 7 7 + User wants to delegate: 8 8 + $ARGUMENTS

+10

modules/opencode/commands/explore.md

reviewed

··· 1 1 + --- 2 2 + description: Explore and load relevant code into context using LSP 3 3 + agent: explore 4 4 + subtask: false 5 5 + --- 6 6 + 7 7 + **Protocol:** Use LSP (goToDefinition, findReferences) to find and read relevant code. Tests included if relevant. End with a 3-bullet summary of what was found. No implementation. 8 8 + 9 9 + User wants to explore: 10 10 + $ARGUMENTS

+3 -8

modules/opencode/commands/grill.md

reviewed

··· 3 3 agent: orchestrator 4 4 --- 5 5 6 6 - Load the grill-me skill and conduct an adversarial review of my plan/decision/design. 7 7 - 8 8 - Walk the full decision tree one question at a time: 6 6 + **Protocol:** Load the grill-me skill and conduct an adversarial review. Walk the full decision tree one question at a time: Surface hidden dependencies and failure modes. Make tradeoffs explicit. Present 2-3 alternative approaches with different tradeoff profiles. Do not proceed to the next question until I have a concrete, defensible answer. 9 7 10 10 - - Surface hidden dependencies and failure modes 11 11 - - Make tradeoffs explicit 12 12 - - Present 2-3 alternative approaches with different tradeoff profiles 13 13 - 14 14 - Do not proceed to the next question until I have a concrete, defensible answer. 8 8 + User wants to grill: 9 9 + $ARGUMENTS