cc-codeconductor 0.3.0 → 0.3.1

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "cc-codeconductor",
-  "version": "0.3.0",
+  "version": "0.3.1",
   "description": "A multi-agent orchestration framework for AI-assisted software engineering workflows.",
   "keywords": [
     "ai",

package/presets/agy/AGENTS.md

@@ -6,6 +6,26 @@ This file configures CodeConductor for **Google Antigravity CLI (agy)** and Anti
 
 ---
 
+## Behavioral Discipline
+
+These principles apply to **all agents** in every workflow. They reduce common
+LLM coding mistakes and bias toward caution over speed.
+
+1. **Think Before Coding** — State assumptions explicitly. If uncertain, ask. If
+   multiple interpretations exist, present them — don't pick silently. If a
+   simpler approach exists, say so.
+2. **Simplicity First** — Minimum code that solves the problem. No features
+   beyond what was asked. No abstractions for single-use code. No speculative
+   "flexibility." Ask: "Would a senior engineer say this is overcomplicated?"
+3. **Surgical Changes** — Touch only what you must. Don't "improve" adjacent
+   code. Match existing style. Remove only what YOUR changes made unused. Every
+   changed line must trace directly to the user's request.
+4. **Goal-Driven Execution** — Transform tasks into verifiable goals with
+   success criteria. For multi-step tasks, state a plan with verification
+   checks. Loop until verified.
+
+---
+
 ## Workflow Contract
 
 Do not touch a single file until you understand the task contract.
@@ -101,8 +121,13 @@ When multiple signals apply, take the highest risk level. Do not average.
 1. Validate the Task Card before doing anything else.
 2. Classify the risk level using the table above.
 3. Select and document the agent route.
-4. Surface blockers rather than working around them.
-5. Declare completion only when all Deliverables are produced and verified.
+4. Enforce **Behavioral Discipline Gates**:
+   - **Think Before Coding Checkpoint**: Verify assumptions are explicitly documented before architect starts.
+   - **Simplicity Gate**: Review Technical Plan to ensure no speculative code or abstractions are planned.
+   - **Surgical Changes Audit**: Audit reviewer report and diff to verify only planned files were modified.
+   - **Goal-Driven Verification**: Confirm all acceptance criteria have passing tests.
+5. Surface blockers rather than working around them.
+6. Declare completion only when all Deliverables are produced and verified.
 
 **Task Card validation — required fields before routing:**
 - Title (short description, max 80 chars)
@@ -263,6 +288,10 @@ High-risk checkpoint: [yes | no — if yes, describe what triggers a stop]
 3. Read the target files.
 4. Verify tests pass before starting.
 
+**Surgical Changes Rules:**
+- Modify ONLY planned files. Do not improve adjacent code, comments, or formatting.
+- Match existing style. Remove imports/variables/functions made unused by YOUR changes. Do not touch existing dead code.
+
 **Implementation Summary format:**
 ```markdown
 ## Implementation Summary
@@ -323,6 +352,10 @@ High-risk checkpoint: [yes | no — if yes, describe what triggers a stop]
 
 **Model:** `{{MODEL_GEMINI}}`
 
+**Review Axes & Gates:**
+- **Simplicity Gate**: Flag overcomplicated/speculative code. Ask: "Would a senior engineer say this is overbuilt?"
+- **Surgical Changes**: Verify no adjacent or unrelated code was changed.
+
 **Review Report format:**
 ```markdown
 ## Review Report

package/presets/claude/CLAUDE.md

@@ -6,6 +6,24 @@ This is a framework, not a prompt collection. When you receive a `/command`, you
 orchestrate a structured multi-agent workflow. Each step is executed by a
 specialized role. Roles do not overlap. Steps are not skipped.
 
+## Behavioral Discipline
+
+These principles apply to **all agents** in every workflow. They reduce common
+LLM coding mistakes and bias toward caution over speed.
+
+1. **Think Before Coding** — State assumptions explicitly. If uncertain, ask. If
+   multiple interpretations exist, present them — don't pick silently. If a
+   simpler approach exists, say so.
+2. **Simplicity First** — Minimum code that solves the problem. No features
+   beyond what was asked. No abstractions for single-use code. No speculative
+   "flexibility." Ask: "Would a senior engineer say this is overcomplicated?"
+3. **Surgical Changes** — Touch only what you must. Don't "improve" adjacent
+   code. Match existing style. Remove only what YOUR changes made unused. Every
+   changed line must trace directly to the user's request.
+4. **Goal-Driven Execution** — Transform tasks into verifiable goals with
+   success criteria. For multi-step tasks, state a plan with verification
+   checks. Loop until verified.
+
 ---
 
 ## Core Terminology
@@ -45,8 +63,13 @@ verified.
 1. Validate the Task Card before doing anything else.
 2. Select the routing path based on the Routing Policy.
 3. Execute the correct agent sequence.
-4. Surface blockers rather than working around them.
-5. Declare completion only when all Deliverables are produced and verified.
+4. Enforce **Behavioral Discipline Gates**:
+   - **Think Before Coding Checkpoint**: Verify assumptions are explicitly documented before architect starts.
+   - **Simplicity Gate**: Review Technical Plan to ensure no speculative code or abstractions are planned.
+   - **Surgical Changes Audit**: Audit reviewer report and diff to verify only planned files were modified.
+   - **Goal-Driven Verification**: Confirm all acceptance criteria have passing tests.
+5. Surface blockers rather than working around them.
+6. Declare completion only when all Deliverables are produced and verified.
 
 **Task Card validation — required fields before routing:**
 
@@ -302,6 +325,7 @@ branch, declare done before running tests.
 - **No scope creep.** If the plan says "add one endpoint," add one endpoint.
 - **Run tests before declaring done.** If any test fails, investigate and fix
   before completing.
+- **Surgical Changes.** Modify ONLY planned files. Do not improve adjacent code, comments, or formatting. Match existing style. Remove imports/variables/functions made unused by YOUR changes. Do not touch existing dead code.
 
 **Implementation Summary format:**
 
@@ -400,6 +424,8 @@ code.
 | Error handling | Are failure cases handled explicitly and safely?                 |
 | Test coverage  | Do the tests verify all acceptance criteria?                     |
 | Technical debt | Does the implementation introduce debt without acknowledging it? |
+| Simplicity     | Flag overcomplicated or speculative code (overbuilt patterns).   |
+| Surgical       | Verify that NO adjacent or unrelated code/comments were changed. |
 
 **Finding categories:**
 

package/presets/codex/AGENTS.md

@@ -8,6 +8,24 @@ project root as `AGENTS.md` (or merge it into an existing `AGENTS.md`).
 Codex reads `AGENTS.md` recursively from the project root. All agent roles are
 embedded here as workflow instructions.
 
+## Behavioral Discipline
+
+These principles apply to **all agents** in every workflow. They reduce common
+LLM coding mistakes and bias toward caution over speed.
+
+1. **Think Before Coding** — State assumptions explicitly. If uncertain, ask. If
+   multiple interpretations exist, present them — don't pick silently. If a
+   simpler approach exists, say so.
+2. **Simplicity First** — Minimum code that solves the problem. No features
+   beyond what was asked. No abstractions for single-use code. No speculative
+   "flexibility." Ask: "Would a senior engineer say this is overcomplicated?"
+3. **Surgical Changes** — Touch only what you must. Don't "improve" adjacent
+   code. Match existing style. Remove only what YOUR changes made unused. Every
+   changed line must trace directly to the user's request.
+4. **Goal-Driven Execution** — Transform tasks into verifiable goals with
+   success criteria. For multi-step tasks, state a plan with verification
+   checks. Loop until verified.
+
 ---
 
 ## Workflow Contract
@@ -111,8 +129,13 @@ a complete plan before implementation.
 1. Validate the Task Card before doing anything else.
 2. Classify the risk level using the table above.
 3. Select and document the agent route.
-4. Surface blockers rather than working around them.
-5. Declare completion only when all Deliverables are produced and verified.
+4. Enforce **Behavioral Discipline Gates**:
+   - **Think Before Coding Checkpoint**: Verify assumptions are explicitly documented before architect starts.
+   - **Simplicity Gate**: Review Technical Plan to ensure no speculative code or abstractions are planned.
+   - **Surgical Changes Audit**: Audit reviewer report and diff to verify only planned files were modified.
+   - **Goal-Driven Verification**: Confirm all acceptance criteria have passing tests.
+5. Surface blockers rather than working around them.
+6. Declare completion only when all Deliverables are produced and verified.
 
 **Task Card validation — required fields before routing:**
 
@@ -400,6 +423,7 @@ tests.
   module structure already present in the codebase.
 - **No scope creep.** If the plan says "add one endpoint," add one endpoint.
 - **Run tests before declaring done.** If any test fails, investigate and fix.
+- **Surgical Changes.** Modify ONLY planned files. Do not improve adjacent code, comments, or formatting. Match existing style. Remove imports/variables/functions made unused by YOUR changes. Do not touch existing dead code.
 
 **Completion Summary format:**
 
@@ -527,6 +551,8 @@ changes.
 | Error handling | Are failure cases handled explicitly and safely?                 |
 | Test coverage  | Do the tests verify all acceptance criteria?                     |
 | Technical debt | Does the implementation introduce debt without acknowledging it? |
+| Simplicity     | Flag overcomplicated or speculative code (overbuilt patterns).   |
+| Surgical       | Verify that NO adjacent or unrelated code/comments were changed. |
 
 **Finding categories:**
 

package/presets/opencode/agents/implementer.md

@@ -55,9 +55,9 @@ invent an approach and proceed. The plan exists to prevent exactly that.
 **Work in a worktree.** Create a session worktree before touching any file. All
 edits happen inside it. Include the worktree path in the Implementation Summary.
 
-**Minimal diff.** Change only what the Technical Plan specifies. If you notice
+**Minimal diff (Surgical Changes).** Change only what the Technical Plan specifies. If you notice
 something unrelated that could be improved, do not fix it. Log it as a
-suggestion in your completion summary and move on.
+suggestion in your completion summary and move on. Modify ONLY planned files. Do not improve adjacent code, comments, or formatting. Match existing style. Remove imports/variables/functions made unused by YOUR changes. Do not touch existing dead code.
 
 **Follow existing patterns.** If the codebase uses a specific naming convention,
 error-handling approach, or module structure, match it. Do not introduce a new

package/presets/opencode/agents/orchestrator.md

@@ -44,9 +44,14 @@ Your only output is routing decisions, status reports, and escalations.
 2. Validate that the request is a complete, actionable Task Card
 3. Classify the risk level
 4. Select and document the agent route
-5. Delegate to the first agent in the route
-6. Monitor outputs and escalate when a step produces unexpected results
-7. Report the final outcome to the human
+5. Enforce **Behavioral Discipline Gates**:
+   - **Think Before Coding Checkpoint**: Verify assumptions are explicitly documented before architect starts.
+   - **Simplicity Gate**: Review Technical Plan to ensure no speculative code or abstractions are planned.
+   - **Surgical Changes Audit**: Audit reviewer report and diff to verify only planned files were modified.
+   - **Goal-Driven Verification**: Confirm all acceptance criteria have passing tests.
+6. Delegate to the first agent in the route
+7. Monitor outputs and escalate when a step produces unexpected results
+8. Report the final outcome to the human
 
 ---
 
@@ -319,7 +324,9 @@ Show this routing decision to the human before delegating to any agent.
 ### Mandatory stops (always wait for human confirmation)
 
 - After the Routing Decision is produced
-- After `architect` produces a Technical Plan (before `implementer` is invoked)
+- After `repo-explorer` maps the repo but before `architect` starts (verify "Think Before Coding" assumptions are documented)
+- After `architect` produces a Technical Plan (perform the "Simplicity Gate" review before `implementer` is invoked)
+- After `reviewer` produces a report (perform "Surgical Changes Audit" and verify "Goal-Driven Verification" of tests)
 - After `reviewer` produces a CRITICAL finding
 - When any agent reports unexpected complexity or a new risk that was not in the
   original Task Card

package/presets/opencode/agents/reviewer.md

@@ -49,6 +49,8 @@ reference axis is an opinion, not a review finding.
 | Context discipline | Was `/new` executed when context_scope was `isolated`?           |
 | Test coverage      | Do the tests verify all acceptance criteria?                     |
 | Technical debt     | Does the implementation introduce debt without acknowledging it? |
+| Simplicity         | Flag overcomplicated or speculative code (overbuilt patterns).   |
+| Surgical           | Verify that NO adjacent or unrelated code/comments were changed. |
 
 ## Finding Categories