Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 3 additions & 9 deletions .agent/aidevops/add-new-mcp-to-aidevops.md
Original file line number Diff line number Diff line change
Expand Up @@ -148,14 +148,14 @@ Before proceeding, ask:
> "Which main agents and subagents should have this MCP enabled?
>
> **Main Agents Available**:
> Plan+, Build+, Accounts, AI-DevOps, Content, Health, Legal, Marketing,
> Build+, Accounts, AI-DevOps, Content, Health, Legal, Marketing,
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This prompt asks about “main agents and subagents”, but the “Main Agents Available” list no longer includes the planning-only option that’s referenced later as @plan-plus, which makes the example a bit inconsistent for enablement decisions. Consider either listing @plan-plus under subagents here or adjusting the wording to match the list.

Fix This in Augment

🤖 Was this useful? React with 👍 or 👎

> Research, Sales, SEO, WordPress
>
> **Recommendation**: Enable globally disabled, then enable per-agent only
> where needed for context efficiency.
>
> **Common patterns**:
> - Codebase/context tools → Plan+, Build+, AI-DevOps, WordPress, Research
> - Codebase/context tools → Build+, AI-DevOps, WordPress, Research
> - Documentation tools → All development agents
> - Domain-specific → Only relevant domain agents
>
Expand All @@ -170,7 +170,7 @@ Record which agents and why:

| Agent | Enabled | Rationale |
|-------|---------|-----------|
| Plan+ | Yes | Needs codebase context for planning |
| @plan-plus | Optional | Subagent for planning-only mode (Build+ handles planning by default) |
| Build+ | Yes | Primary development agent |
| AI-DevOps | Yes | Infrastructure development |
| WordPress | No | Not relevant to WordPress tasks |
Expand Down Expand Up @@ -246,12 +246,6 @@ Edit `.agent/scripts/generate-opencode-agents.sh`:
Based on the agent enablement decision from Step 2:

```python
"Plan+": {
"tools": {
# ... existing tools ...
"{mcp-name}_*": True # Only if enabled for this agent
}
},
"Build+": {
"tools": {
# ... existing tools ...
Expand Down
2 changes: 1 addition & 1 deletion .agent/aidevops/troubleshooting.md
Original file line number Diff line number Diff line change
Expand Up @@ -248,7 +248,7 @@ opencode run "/new-command arg1" --agent Build+

```bash
~/.aidevops/agents/scripts/opencode-test-helper.sh test-mcp dataforseo SEO
~/.aidevops/agents/scripts/opencode-test-helper.sh test-agent Plan+
~/.aidevops/agents/scripts/opencode-test-helper.sh test-agent Build+
```

## Performance Optimization
Expand Down
2 changes: 1 addition & 1 deletion .agent/scripts/commands/list-todo.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
---
description: List tasks from TODO.md with sorting and filtering options
agent: Plan+
agent: Build+
mode: subagent
---

Expand Down
2 changes: 1 addition & 1 deletion .agent/scripts/commands/save-todo.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
---
description: Save current discussion as task or plan (auto-detects complexity)
agent: Plan+
agent: Build+
mode: subagent
---

Expand Down
2 changes: 1 addition & 1 deletion .agent/scripts/commands/show-plan.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
---
description: Show detailed information about a specific plan from PLANS.md
agent: Plan+
agent: Build+
mode: subagent
---

Expand Down
2 changes: 1 addition & 1 deletion .agent/tools/context/context-guardrails.md
Original file line number Diff line number Diff line change
Expand Up @@ -160,7 +160,7 @@ Before attempting edits, verify you have the required tools:
```text
Self-check: "Do I have Edit/Write/Bash tools for this task?"

If NO (e.g., in Plan+ agent):
If NO (e.g., in @plan-plus subagent or read-only mode):
-> Suggest: "This task requires edits. Please switch to Build+ agent."

If YES:
Expand Down
12 changes: 6 additions & 6 deletions .agent/tools/opencode/opencode.md
Original file line number Diff line number Diff line change
Expand Up @@ -358,7 +358,7 @@ opencode run --attach http://localhost:4096 "Another test" --agent SEO
|----------|---------|
| New MCP added | `opencode run "List tools from [mcp]_*" --agent [agent]` |
| MCP auth issues | `opencode run "Call [mcp]_[tool]" --agent [agent] 2>&1` |
| Agent permissions | `opencode run "Try to write a file" --agent Plan+` |
| Agent permissions | `opencode run "Try to write a file" --agent Build+` |
| Slash command | `opencode run "/new-command arg1 arg2" --agent Build+` |
| Quick task | `opencode run "Do X quickly" --agent Build+` |

Expand All @@ -371,7 +371,7 @@ Use the helper script for common testing tasks:
~/.aidevops/agents/scripts/opencode-test-helper.sh test-mcp dataforseo SEO

# Test agent permissions
~/.aidevops/agents/scripts/opencode-test-helper.sh test-agent Plan+
~/.aidevops/agents/scripts/opencode-test-helper.sh test-agent Build+
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This doc now shows opencode-test-helper.sh test-agent Build+, but the helper script’s own show_help() examples still reference test-agent Plan+, which can lead to conflicting guidance. Consider aligning the examples so users aren’t unsure which agent name is expected.

Other Locations
  • .agent/aidevops/troubleshooting.md:251

Fix This in Augment

🤖 Was this useful? React with 👍 or 👎


# List tools available to agent
~/.aidevops/agents/scripts/opencode-test-helper.sh list-tools Build+
Expand Down Expand Up @@ -532,7 +532,7 @@ When a parent agent uses the `task` tool to spawn a subagent:
| `write: false, edit: false, bash: true` | NO - bash can write files |
| `write: false, edit: false, bash: false, task: false` | YES - truly read-only |

**Example**: Plan+ with `task: true` could call a subagent that creates files, defeating its read-only purpose.
**Example**: A read-only agent (like @plan-plus) with `task: true` could call a subagent that creates files, defeating its read-only purpose.

### Bash Escapes All Restrictions

Expand All @@ -543,12 +543,12 @@ When `bash: true`, the agent can execute ANY shell command, including:

**For true read-only behavior**: Set both `bash: false` AND `task: false`

### Plan+ Read-Only Configuration
### @plan-plus Read-Only Configuration

Plan+ is configured as strictly read-only:
The @plan-plus subagent is configured as strictly read-only:

```json
"Plan+": {
"@plan-plus": {
"permission": {
"edit": "deny",
"write": "deny",
Expand Down
4 changes: 2 additions & 2 deletions .agent/workflows/conversation-starter.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ mode: subagent
---
# Conversation Starter Prompts

Shared prompts for Plan+ and Build+ agents to ensure consistent UX.
Shared prompts for Build+ agent to ensure consistent UX.

## Inside Git Repository

Expand All @@ -22,7 +22,7 @@ If on `main` branch, include this note in the prompt:

What are you working on?

**Planning & Analysis** (Plan+):
**Planning & Analysis** (Build+ deliberation mode):
>
> 1. Architecture Analysis
> 2. Code Review (`workflows/code-audit-remote.md`)
Expand Down
4 changes: 2 additions & 2 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -91,7 +91,7 @@ The result: AI agents that work *with* your development process, not around it.

### Agent Structure

- 15 primary agents (Plan+, Build+, SEO, Marketing, etc.)
- Primary agents (Build+, SEO, Marketing, etc.) with @plan-plus subagent for planning-only mode
- 552+ subagent markdown files organized by domain
- 147 helper scripts in `.agent/scripts/`
- 14 slash commands for common workflows
Expand Down Expand Up @@ -1238,7 +1238,7 @@ Ordered as they appear in OpenCode Tab selector and other AI assistants (15 tota

| Name | File | Purpose | MCPs Enabled |
|------|------|---------|--------------|
| Plan+ | `plan-plus.md` | Planning with semantic search, writes to TODO.md/todo/ | context7, augment, repomix |
| @plan-plus | `plan-plus.md` | Planning-only subagent (Build+ handles planning by default) | context7, augment, repomix |
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In the “Main Agents” table, the @plan-plus row is described as a “subagent”, which conflicts with the section label (and the earlier “primary agents … with @plan-plus subagent” line). Consider clarifying whether @plan-plus is meant to be tab-selectable as a main agent or only @mentionable.

Fix This in Augment

🤖 Was this useful? React with 👍 or 👎

| Build+ | `build-plus.md` | Enhanced Build with context tools | context7, augment, repomix |
| Build-Agent | `build-agent.md` | Design and improve AI agents | context7, augment, repomix |
| Build-MCP | `build-mcp.md` | Build MCP servers with TS+Bun+ElysiaJS | context7, augment, repomix |
Expand Down