cc-codeconductor 1.4.2 → 1.5.1

package/presets/opencode/README.md

@@ -2,8 +2,8 @@
 
 ## Model Selection Guide
 
-This preset supports both **Claude (Anthropic)** and **OpenCode Go (Chinese
-providers)** models.
+This preset uses the four requested **OpenCode Go** models. Claude (Anthropic)
+can be selected as an explicit override.
 
 ---
 
@@ -13,37 +13,39 @@ providers)** models.
 
 | Model                       | Strength                        | Best For                        |
 | --------------------------- | ------------------------------- | ------------------------------- |
-| `claude-opus-4-7`           | Complex reasoning, architecture | Architect, complex design       |
-| `claude-sonnet-4-6`         | Balanced, general purpose       | Default for most agents         |
+| `claude-opus-5`             | Complex reasoning, architecture | Architect, complex design       |
+| `claude-sonnet-5`           | Balanced, general purpose       | Implementation, testing         |
 | `claude-haiku-4-5-20251001` | Fast, lightweight               | Task Coach, Docs, Repo Explorer |
 
 ### OpenCode Go models
 
-Canonical slugs live in `src/presets/models/*.yml`. Production roles never use
-`muse-spark-1.2-contributor` (trains on prompts). `ox-alpha-free` is an optional
-fallback, not a default.
+Canonical role assignments live in `src/presets/models/roles.yml`. Model IDs
+were verified against [OpenCode Go](https://opencode.ai/docs/go/) and the live
+catalog on September 25, 2026.
 
 | Model | Best for |
 | --- | --- |
-| `deepseek-v4-pro` | Architect |
-| `kimi-k3` | Security reviewer |
-| `qwen3.8-max` | Reviewer, contract-builder |
-| `glm-5.3` | Devil, complexity-auditor |
-| `mimo-v2.5` | Implementer |
-| `minimax-m3` | Tester |
-| `qwen3.7-plus` | Orchestrator (TUI default) |
-| `gpt-5.6-luna` | Task coach, planner |
-| `deepseek-v4-flash` | Goal planner |
-| `longcat-2.0` | Repo explorer |
-| `hy3` | Docs |
+| `opencode-go/deepseek-v4.1-flash` | Architect, reviewer, contract-builder, orchestrator (TUI default), planner, goal-planner |
+| `opencode-go/mimo-v2.6-flash` | Implementer, tester |
+| `opencode-go/glm-5.3-flash` | Security-reviewer, devil, complexity-auditor, task-coach, repo-explorer |
+| `opencode-go/muse-spark-1.3-contributor` | Docs |
+
+These assignments are workflow choices, not a provider ranking. Generated
+agents leave sampling and reasoning settings at provider defaults rather than
+copying Claude-specific `effort` or forcing a shared `temperature` onto different
+models. No unverified variants are configured.
+
+Muse Contributor is included at the request of the maintainer. OpenCode documents
+that it uses prompts/completions for training, is not ZDR, and has regional
+restrictions. Do not send confidential material to that model; override the docs
+agent with another verified model when required.
 
 ---
 
 ## Agent Model Matrix
 
-See `skills/cc-update-preset-models/references/role-map.md`. OpenCode column
-examples: architect `deepseek-v4-pro`, implementer `mimo-v2.5`, tester
-`minimax-m3`, reviewer `qwen3.8-max`.
+See `skills/cc-update-preset-models/references/role-map.md` for the complete
+cross-preset role matrix.
 
 ---
 
@@ -66,6 +68,10 @@ examples: architect `deepseek-v4-pro`, implementer `mimo-v2.5`, tester
 
 This preset uses OpenCode's permission system (v1.1.1+) with granular control:
 
+Keep the V1-compatible `permission` dictionary for the `opencode` CLI. V2's
+`permissions` rule list is exclusive to `opencode2` and is rejected by OpenCode
+V1. Validate generated files with the actual CLI version before migrating formats.
+
 - **Global defaults**: Most operations require approval (`ask`)
 - **Read access**: Allowed by default, with sensitive files denied
 - **Write/Edit**: Requires approval, with protected paths denied
@@ -87,8 +93,8 @@ The following paths are denied by default:
 
 ### Simple Tasks (Q&A, intake, documentation)
 
-**Recommended:** `qwen3.6-plus` (OpenCode Go) or `claude-haiku-4-5-20251001`
-(Claude)
+**Preset choice:** `glm-5.3-flash` for intake/exploration;
+`muse-spark-1.3-contributor` for non-confidential docs.
 
 - Task Coach intake
 - Repo Explorer mapping
@@ -96,8 +102,8 @@ The following paths are denied by default:
 
 ### Medium Tasks (Implementation, testing)
 
-**Recommended:** `mimo-v2.5` or `minimax-m3` (OpenCode Go) or
-`claude-sonnet-4-6` (Claude)
+**Preset choice:** `mimo-v2.6-flash` for implementation and testing;
+`deepseek-v4.1-flash` for reviews.
 
 - Implementer code writing
 - Tester test generation
@@ -105,7 +111,8 @@ The following paths are denied by default:
 
 ### Complex Tasks (Architecture, security, multi-agent coordination)
 
-**Recommended:** `deepseek-v4-pro` (OpenCode Go) or `claude-opus-4-7` (Claude)
+**Preset choice:** `deepseek-v4.1-flash` for design/coordination;
+`glm-5.3-flash` for defensive security review.
 
 - Architect technical design
 - Orchestrator routing decisions
@@ -115,17 +122,14 @@ The following paths are denied by default:
 
 ## Usage in Agent Contracts
 
-Each agent file in `agents/` contains a model selection table in its
-frontmatter:
+Agent templates contain placeholders. Installation resolves the role model
+into YAML frontmatter:
 
 ```yaml
 ---
 description: ...
-# Model Selection
-| Provider | Model | Use Case |
-|----------|-------|----------|
-| Claude | claude-opus-4-7 | Complex architecture, ADRs |
-| OpenCode Go | deepseek-v4-pro | Best — excels at reasoning |
+model: opencode-go/deepseek-v4.1-flash
+mode: subagent
 ---
 ```
 
@@ -144,21 +148,10 @@ configuration in `opencode.jsonc`.
 
 ## Environment Variables
 
-OpenCode Go requires appropriate API keys. Set these in your environment:
-
-```bash
-# DeepSeek
-export DEEPSEEK_API_KEY="your-key"
-
-# Qwen (Alibaba)
-export DASHSCOPE_API_KEY="your-key"
-
-# MiniMax
-export MINIMAX_API_KEY="your-key"
-
-# Kimi (Moonshot)
-export KIMI_API_KEY="your-key"
-```
+Connect to **OpenCode Go** through `/connect` and use the Go API key. The
+`opencode-go/` provider does not use individual DeepSeek, GLM, MiMo, or Meta keys.
+Do not store API keys in the preset. Run `opencode models opencode-go` to check
+availability before using or overriding a role model.
 
 ---
 
@@ -166,10 +159,10 @@ export KIMI_API_KEY="your-key"
 
 | Scenario                        | Recommended                                        |
 | ------------------------------- | -------------------------------------------------- |
-| Complex reasoning, architecture | OpenCode Go (`deepseek-v4-pro`) or Claude (`opus`) |
-| Fast iteration, simple tasks    | OpenCode Go (`qwen3.6-plus`) or Claude (`haiku`)   |
-| Code implementation             | OpenCode Go (`mimo-v2.5`) or Claude (`sonnet`) |
-| Budget constraints              | OpenCode Go (generally lower cost)                 |
+| Complex reasoning, architecture | OpenCode Go (`deepseek-v4.1-flash`) or Claude (`opus`) |
+| Fast iteration, simple tasks    | OpenCode Go (`glm-5.3-flash`) or Claude (`haiku`)   |
+| Code implementation             | OpenCode Go (`mimo-v2.6-flash`) or Claude (`sonnet`) |
+| Non-confidential documentation   | OpenCode Go (`muse-spark-1.3-contributor`) |
 | Availability issues             | Switch to alternative from the matrix              |
 
 ## Approach

package/presets/opencode/commands/cc-api-contract.md

@@ -12,11 +12,10 @@ API contract request: $ARGUMENTS
 
 Command: `api-contract` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command api-contract "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command api-contract "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile api-contract --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command api-contract --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile api-contract --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command api-contract --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command api-contract --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-backlog.md

@@ -17,11 +17,10 @@ deliver implementation — that is `/cc-openspec`.
 
 Command: `backlog` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command backlog "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command backlog "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile backlog --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command backlog --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile backlog --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command backlog --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command backlog --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-clarify.md

@@ -10,11 +10,10 @@ Clarify request: $ARGUMENTS
 
 Command: `clarify` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command clarify "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command clarify "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile clarify --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command clarify --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile clarify --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command clarify --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command clarify --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-council.md

@@ -11,11 +11,10 @@ Task request: $ARGUMENTS
 Command: `council` (fixed for this workflow — do not infer from user text)
 command: council
 
-1. Run: `npx cc-codeconductor ccep parse --command council "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command council "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile council --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command council --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile council --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command council --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command council --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-db-migration.md

@@ -12,11 +12,10 @@ Migration request: $ARGUMENTS
 
 Command: `db-migration` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command db-migration "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command db-migration "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile db-migration --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command db-migration --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile db-migration --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command db-migration --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command db-migration --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---

package/presets/opencode/commands/cc-explore.md

@@ -10,11 +10,10 @@ Explore request: $ARGUMENTS
 
 Command: `explore` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command explore "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command explore "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile explore --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command explore --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile explore --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command explore --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command explore --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-feature.md

@@ -34,11 +34,10 @@ unavailable visual or emulator checks remain pending rather than claimed complet
 
 Command: `feature` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command feature "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command feature "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile feature --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command feature --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile feature --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command feature --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command feature --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---

package/presets/opencode/commands/cc-fix.md

@@ -44,11 +44,10 @@ unavailable visual or emulator checks remain pending rather than claimed complet
 
 Command: `fix` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command fix "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command fix "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile fix --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command fix --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile fix --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command fix --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command fix --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---

package/presets/opencode/commands/cc-handoff.md

@@ -10,11 +10,10 @@ Handoff request: $ARGUMENTS
 
 Command: `handoff` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command handoff "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command handoff "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile handoff --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command handoff --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile handoff --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command handoff --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command handoff --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 
@@ -34,6 +33,10 @@ pass/fail (not log dumps), open questions, and the next `/cc-` command.
 Link the Delivery Ledger and verification evidence when they exist; do not
 repeat the original request or transcript. Do not edit source or tests.
 
+If a Delivery Ledger exists, run `npx cc-codeconductor odd handoff --id <ledger-id> --output json`
+and use its `handoff` envelope as the source for these fields. Secret redaction
+still applies.
+
 ---
 
 ## Completion

package/presets/opencode/commands/cc-iterative.md

@@ -12,11 +12,10 @@ Iterative request: $ARGUMENTS
 
 Command: `iterative` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command iterative "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command iterative "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile iterative --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command iterative --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile iterative --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command iterative --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command iterative --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---

package/presets/opencode/commands/cc-odd.md

@@ -7,9 +7,10 @@ description: Create or resume an opt-in Delivery Ledger for substantial authoriz
 ## Step 0 — CCEP Bootstrap
 
 ```bash
-bun run dev ccep parse --command odd "$ARGUMENTS"
-bun run dev ccep resolve --command odd "$ARGUMENTS"
-bun run dev ccep profile --command odd
+bun run dev ccep profile odd --output json
+bun run dev ccep compile --command odd --phase <phase-id> "$ARGUMENTS" --view prompt --output json
 ```
 
+Pass each subagent only the compiled `prompt` for its phase.
+
 Create a ledger only after authorization and tracked coordination. Read-only and small work do not create state.

package/presets/opencode/commands/cc-openspec.md

@@ -32,7 +32,7 @@ stays with the general `android` skill. Audit-only requests remain read-only, an
 unavailable visual or emulator checks remain pending rather than claimed complete.
 
 
-## Step 0 — Validate (mandatory gate)
+## Step 0a — Validate (mandatory gate)
 
 Run `npx cc-codeconductor openspec validate`. If invalid, show errors and recommendations, then **STOP**.
 
@@ -42,11 +42,10 @@ Run `npx cc-codeconductor openspec validate`. If invalid, show errors and recomm
 
 Command: `openspec` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command openspec "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command openspec "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile openspec --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command openspec --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile openspec --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command openspec --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command openspec --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---
@@ -95,7 +94,7 @@ For each pending card: `npx cc-codeconductor openspec next`, then invoke the lis
 - implement → `implementer`
 - review → `reviewer`
 
-Use isolated context (`/clear` between phases). Implementer uses a git worktree.
+Run each phase as a subagent with isolated context. Implementer uses a git worktree.
 
 ---
 

package/presets/opencode/commands/cc-pagespeed.md

@@ -14,11 +14,10 @@ Audit URL: $ARGUMENTS
 
 Command: `pagespeed` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command pagespeed "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command pagespeed "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile pagespeed --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command pagespeed --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile pagespeed --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command pagespeed --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command pagespeed --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 
@@ -39,9 +38,8 @@ without a key, but CrUX field data requires it.
 Call the PageSpeed Insights API for the URL provided above.
 
 - Strategy: `both` (mobile + desktop) unless the user specified otherwise.
-- If Bun is available, use `~/.claude/skills/pagespeed-perf/scripts/run.ts`.
-- If Bun is unavailable, call PSI directly via WebFetch and fetch the HTML
-  source for manual resource-hint analysis.
+- Call PSI directly via WebFetch and fetch the HTML source for manual
+  resource-hint analysis.
 
 ---
 

package/presets/opencode/commands/cc-prototype.md

@@ -10,11 +10,10 @@ Prototype request: $ARGUMENTS
 
 Command: `prototype` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command prototype "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command prototype "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile prototype --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command prototype --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile prototype --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command prototype --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command prototype --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-refactor.md

@@ -37,11 +37,10 @@ If coverage is insufficient:
 
 Command: `refactor` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command refactor "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command refactor "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile refactor --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command refactor --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile refactor --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command refactor --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command refactor --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-review.md

@@ -43,11 +43,10 @@ unavailable visual or emulator checks remain pending rather than claimed complet
 
 Command: `review` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command review "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command review "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile review --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command review --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile review --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command review --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command review --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 
@@ -118,6 +117,37 @@ reviewer must evaluate the diff against the following checklist:
 
 ---
 
+## Severity Rubric
+
+Classify every finding using these concrete anchors — not general judgment —
+so the same finding gets the same category on any run.
+
+**CRITICAL** (always):
+
+- Unauthenticated access to a protected resource
+- Injection (SQL, command, XSS) reachable from untrusted input
+- A secret or credential present in the diff
+- A destructive operation with no confirmation
+- The build or test suite is broken
+- A direct contradiction of an explicit acceptance criterion
+
+**WARNING**:
+
+- Missing test for a new branch or edge case
+- An error is silently swallowed — no log, no propagation
+- A change outside the declared scope
+- Inconsistent with an established pattern already in the module
+- Missing input validation in an internal function (not exposed at a trust boundary)
+
+**SUGGESTION**:
+
+- Style or naming
+- Minor readability
+- Speculative refactor that wasn't requested
+- Non-blocking micro-optimization
+
+---
+
 ## Step 3 — Review Report
 
 reviewer produces a structured Review Report with findings in three categories:

package/presets/opencode/commands/cc-scorecard.md

@@ -11,11 +11,10 @@ Scope: $ARGUMENTS
 
 Command: `scorecard` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command scorecard "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command scorecard "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile scorecard --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command scorecard --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile scorecard --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command scorecard --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command scorecard --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
 
 ---
 

package/presets/opencode/commands/cc-security.md

@@ -47,11 +47,10 @@ payloads, malware, or attack procedures.
 
 Command: `security` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command security "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command security "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile security --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command security --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile security --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command security --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command security --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---
@@ -62,7 +61,7 @@ If `graphify-out/graph.json` exists, run `graphify query "$ARGUMENTS"` (and
 `graphify path` / `graphify explain` when needed). Then invoke `repo-explorer`
 to map modules, conventions, and impact radius. Do not write code in this step.
 Record a Repo Map artifact before intake. Load the matching `security-*` skill
-for the named domain (see `.claude/skills/security-*/SKILL.md`). Keep the OWASP
+for the named domain (see `.opencode/skills/security-*/SKILL.md`). Keep the OWASP
 `security` skill for application-security reviews.
 
 ---

package/presets/opencode/commands/cc-spec-mutation.md

@@ -22,11 +22,10 @@ Describe what behavior you want to implement. Include:
 
 Command: `spec-mutation` (fixed for this workflow — do not infer from user text)
 
-1. Run: `npx cc-codeconductor ccep parse --command spec-mutation "$ARGUMENTS" --output json`
-2. Run: `npx cc-codeconductor ccep resolve --command spec-mutation "$ARGUMENTS" --output json`
-3. Run: `npx cc-codeconductor ccep profile spec-mutation --output json`
-4. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command spec-mutation --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
-5. Delegate to subagents using compiled CCEP prompts — never forward raw `$ARGUMENTS` to planners.
+1. Run: `npx cc-codeconductor ccep profile spec-mutation --output json` to get the phases and their roles.
+2. For each delegated phase, run: `npx cc-codeconductor ccep compile --command spec-mutation --phase <phase-id> "$ARGUMENTS" --view prompt --output json`
+3. After planner/intake JSON is available, run: `npx cc-codeconductor ccep evaluate --command spec-mutation --input <planner.json> --output json`. If `stop` is true, show questions or risks and wait for human input.
+4. Pass each subagent only the compiled `prompt` for its phase — never forward raw `$ARGUMENTS` to planners.
    Canonical delivery order is test-before-implement whenever both phases apply.
 
 ---