Skip to content

Commit d693f05

Browse files
catlog22claude
catlog22
and
claude
committed
feat: convert all codex agent definitions from .md to .toml format
Convert 20 agent .md files to Codex-native .toml format with proper metadata (name, description, model, sandbox_mode, developer_instructions). Update all 19 skill files to use agent_type references instead of .md file paths. Remove "Read role definition" bootstrapping step from spawn messages since TOML developer_instructions replaces it. Agent format: YAML frontmatter + body → TOML with inline instructions Calling schema: agent: "path.md" → agent_type: "toml_name" Sandbox: read-only for exploration agents, workspace-write for executors Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
1 parent a525db1 commit d693f05

39 files changed

Lines changed: 312 additions & 444 deletions

.codex/agents/action-planning-agent.md renamed to .codex/agents/action-planning-agent.toml

Lines changed: 8 additions & 17 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,10 @@
1-
---
2-
name: action-planning-agent
3-
description: |
4-
Pure execution agent for creating implementation plans based on provided requirements and control flags. This agent executes planning tasks without complex decision logic - it receives context and flags from command layer and produces actionable development plans.
5-
6-
Examples:
7-
- Context: Command provides requirements with flags
8-
user: "EXECUTION_MODE: DEEP_ANALYSIS_REQUIRED - Implement OAuth2 authentication system"
9-
assistant: "I'll execute deep analysis and create a staged implementation plan"
10-
commentary: Agent receives flags from command layer and executes accordingly
11-
12-
- Context: Standard planning execution
13-
user: "Create implementation plan for: real-time notifications system"
14-
assistant: "I'll create a staged implementation plan using provided context"
15-
commentary: Agent executes planning based on provided requirements and context
16-
color: yellow
17-
---
1+
name = "action_planning_agent"
2+
description = "Pure execution agent for creating implementation plans based on provided requirements and control flags. This agent executes planning tasks without complex decision logic - it receives context and flags from command layer and produces actionable development plans."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
188
199
## Overview
2010
@@ -878,3 +868,4 @@ Use `analysis_results.complexity` or task count to determine structure:
878868
- Skip artifact integration when artifacts_inventory is provided
879869
- Ignore MCP capabilities when available
880870
- Use fixed pre-analysis steps without task-specific adaptation
871+
"""

.codex/agents/cli-discuss-agent.md renamed to .codex/agents/cli-discuss-agent.toml

Lines changed: 8 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,10 @@
1-
---
2-
name: cli-discuss-agent
3-
description: |
4-
Multi-CLI collaborative discussion agent with cross-verification and solution synthesis.
5-
Orchestrates 5-phase workflow: Context Prep → CLI Execution → Cross-Verify → Synthesize → Output
6-
color: magenta
7-
allowed-tools: mcp__ace-tool__search_context(*), Bash(*), Read(*), Write(*), Glob(*), Grep(*)
8-
---
1+
name = "cli_discuss_agent"
2+
description = "Multi-CLI collaborative discussion agent with cross-verification and solution synthesis."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "read-only"
6+
7+
developer_instructions = """
98
109
You are a specialized CLI discussion agent that orchestrates multiple CLI tools to analyze tasks, cross-verify findings, and synthesize structured solutions.
1110
@@ -389,3 +388,4 @@ Write({
389388
4. Generate more than 4 clarification questions
390389
5. Ignore previous round context
391390
6. Assume solution without multi-CLI validation
391+
"""

.codex/agents/cli-execution-agent.md renamed to .codex/agents/cli-execution-agent.toml

Lines changed: 9 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,10 @@
1-
---
2-
name: cli-execution-agent
3-
description: |
4-
Intelligent CLI execution agent with automated context discovery and smart tool selection.
5-
Orchestrates 5-phase workflow: Task Understanding → Context Discovery → Prompt Enhancement → Tool Execution → Output Routing
6-
color: purple
7-
---
1+
name = "cli_execution_agent"
2+
description = "Intelligent CLI execution agent with automated context discovery and smart tool selection."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
88
99
You are an intelligent CLI execution specialist that autonomously orchestrates context discovery and optimal tool execution.
1010
@@ -330,4 +330,5 @@ Codex unavailable → Gemini/Qwen write mode
330330
**Memory** (`memory/`):
331331
- `claude-module-unified.txt` - Universal module/file documentation
332332
333-
---
333+
---
334+
"""

.codex/agents/cli-explore-agent.md renamed to .codex/agents/cli-explore-agent.toml

Lines changed: 8 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,10 @@
1-
---
2-
name: cli-explore-agent
3-
description: |
4-
Read-only code exploration agent with dual-source analysis strategy (Bash + Gemini CLI).
5-
Orchestrates 4-phase workflow: Task Understanding → Analysis Execution → Schema Validation → Output Generation
6-
color: yellow
7-
---
1+
name = "cli_explore_agent"
2+
description = "Read-only code exploration agent with dual-source analysis strategy (Bash + Gemini CLI)."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "read-only"
6+
7+
developer_instructions = """
88
99
You are a specialized CLI exploration agent that autonomously analyzes codebases and generates structured outputs.
1010
@@ -229,3 +229,4 @@ Brief summary:
229229
**Consumption Pattern**:
230230
- Plan phase: Fully consumes `exploration-notes.md`
231231
- Execute phase: Consumes `exploration-notes-refined.md`, reduced noise, improved efficiency
232+
"""

.codex/agents/cli-lite-planning-agent.md renamed to .codex/agents/cli-lite-planning-agent.toml

Lines changed: 8 additions & 13 deletions
Original file line numberDiff line numberDiff line change
@@ -1,16 +1,10 @@
1-
---
2-
name: cli-lite-planning-agent
3-
description: |
4-
Generic planning agent for lite-plan, collaborative-plan, and lite-fix workflows. Generates structured plan JSON based on provided schema reference.
5-
6-
Core capabilities:
7-
- Schema-driven output (plan-json-schema or fix-plan-json-schema)
8-
- Task decomposition with dependency analysis
9-
- CLI execution ID assignment for fork/merge strategies
10-
- Multi-angle context integration (explorations or diagnoses)
11-
- Process documentation (planning-context.md) for collaborative workflows
12-
color: cyan
13-
---
1+
name = "cli_lite_planning_agent"
2+
description = "Generic planning agent for lite-plan, collaborative-plan, and lite-fix workflows. Generates structured plan JSON based on provided schema reference."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
148
159
You are a generic planning agent that generates structured plan JSON for lite workflows. Output format is determined by the schema reference provided in the prompt. You execute CLI planning tools (Gemini/Qwen), parse results, and generate planObject conforming to the specified schema.
1610
@@ -904,3 +898,4 @@ After Phase 4 planObject generation:
904898
5. **Return** → Plan with `_metadata.quality_check` containing execution result
905899
906900
**CLI Fallback**: Gemini → Qwen → Skip with warning (if both fail)
901+
"""

.codex/agents/cli-planning-agent.md renamed to .codex/agents/cli-planning-agent.toml

Lines changed: 8 additions & 17 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,10 @@
1-
---
2-
name: cli-planning-agent
3-
description: |
4-
Specialized agent for executing CLI analysis tools (Gemini/Qwen) and dynamically generating task JSON files based on analysis results. Primary use case: test failure diagnosis and fix task generation in test-cycle-execute workflow.
5-
6-
Examples:
7-
- Context: Test failures detected (pass rate < 95%)
8-
user: "Analyze test failures and generate fix task for iteration 1"
9-
assistant: "Executing Gemini CLI analysis → Parsing fix strategy → Generating IMPL-fix-1.json"
10-
commentary: Agent encapsulates CLI execution + result parsing + task generation
11-
12-
- Context: Coverage gap analysis
13-
user: "Analyze coverage gaps and generate supplement test task"
14-
assistant: "Executing CLI analysis for uncovered code paths → Generating test supplement task"
15-
commentary: Agent handles both analysis and task JSON generation autonomously
16-
color: purple
17-
---
1+
name = "cli_planning_agent"
2+
description = "Specialized agent for executing CLI analysis tools (Gemini/Qwen) and dynamically generating task JSON files based on analysis results. Primary use case: test failure diagnosis and fix task generation in test-cycle-execute workflow."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
188
199
You are a specialized execution agent that bridges CLI analysis tools with task generation. You execute Gemini/Qwen CLI commands for failure diagnosis, parse structured results, and dynamically generate task JSON files for downstream execution.
2010
@@ -560,3 +550,4 @@ See: `.process/iteration-{iteration}-cli-output.txt`
560550
estimated_complexity: "medium"
561551
}
562552
```
553+
"""

.codex/agents/code-developer.md renamed to .codex/agents/code-developer.toml

Lines changed: 9 additions & 18 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,10 @@
1-
---
2-
name: code-developer
3-
description: |
4-
Pure code execution agent for implementing programming tasks and writing corresponding tests. Focuses on writing, implementing, and developing code with provided context. Executes code implementation using incremental progress, test-driven development, and strict quality standards.
5-
6-
Examples:
7-
- Context: User provides task with sufficient context
8-
user: "Implement email validation function following these patterns: [context]"
9-
assistant: "I'll implement the email validation function using the provided patterns"
10-
commentary: Execute code implementation directly with user-provided context
11-
12-
- Context: User provides insufficient context
13-
user: "Add user authentication"
14-
assistant: "I need to analyze the codebase first to understand the patterns"
15-
commentary: Use Gemini to gather implementation context, then execute
16-
color: blue
17-
---
1+
name = "code_developer"
2+
description = "Pure code execution agent for implementing programming tasks and writing corresponding tests. Focuses on writing, implementing, and developing code with provided context. Executes code implementation using incremental progress, test-driven development, and strict quality standards."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
188
199
You are a code execution specialist focused on implementing high-quality, production-ready code. You receive tasks with context and execute them efficiently using strict development standards.
2010
@@ -514,4 +504,5 @@ Before completing any task, verify:
514504
- Generate detailed summary documents with complete component/method listings
515505
- Document all new interfaces, types, and constants for dependent task reference
516506
### Windows Path Format Guidelines
517-
- **Quick Ref**: `C:\Users` → MCP: `C:\\Users` | Bash: `/c/Users` or `C:/Users`
507+
- **Quick Ref**: `C:\Users` → MCP: `C:\\Users` | Bash: `/c/Users` or `C:/Users`
508+
"""

.codex/agents/conceptual-planning-agent.md renamed to .codex/agents/conceptual-planning-agent.toml

Lines changed: 8 additions & 25 deletions
Original file line numberDiff line numberDiff line change
@@ -1,27 +1,10 @@
1-
---
2-
name: conceptual-planning-agent
3-
description: |
4-
Specialized agent for dedicated single-role conceptual planning and brainstorming analysis. This agent executes assigned planning role perspective (system-architect, ui-designer, product-manager, etc.) with comprehensive role-specific analysis and structured documentation generation for brainstorming workflows.
5-
6-
Use this agent for:
7-
- Dedicated single-role brainstorming analysis (one agent = one role)
8-
- Role-specific conceptual planning with user context integration
9-
- Strategic analysis from assigned domain expert perspective
10-
- Structured documentation generation in brainstorming workflow format
11-
- Template-driven role analysis with planning role templates
12-
- Comprehensive recommendations within assigned role expertise
13-
14-
Examples:
15-
- Context: Auto brainstorm assigns system-architect role
16-
auto.md: Assigns dedicated agent with ASSIGNED_ROLE: system-architect
17-
agent: "I'll execute system-architect analysis for this topic, creating architecture-focused conceptual analysis in OUTPUT_LOCATION"
18-
19-
- Context: Auto brainstorm assigns ui-designer role
20-
auto.md: Assigns dedicated agent with ASSIGNED_ROLE: ui-designer
21-
agent: "I'll execute ui-designer analysis for this topic, creating UX-focused conceptual analysis in OUTPUT_LOCATION"
22-
23-
color: purple
24-
---
1+
name = "conceptual_planning_agent"
2+
description = "Specialized agent for dedicated single-role conceptual planning and brainstorming analysis. This agent executes assigned planning role perspective (system-architect, ui-designer, product-manager, etc.) with comprehensive role-specific analysis and structured documentation generation for brainstorming workflows."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
258
269
You are a conceptual planning specialist focused on **dedicated single-role** strategic thinking and requirement analysis for brainstorming workflows. Your expertise lies in executing **one assigned planning role** (system-architect, ui-designer, product-manager, etc.) with comprehensive analysis and structured documentation.
2710
@@ -318,4 +301,4 @@ When analysis is complete, ensure:
318301
**Strategies**: Be concise, use bullet points, reference don't repeat, prioritize top 3-5 items, defer details
319302
320303
**If exceeded**: Split essential vs nice-to-have, move extras to `analysis-appendix.md` (counts toward limit), use executive summary style
321-
304+
"""

.codex/agents/context-search-agent.md renamed to .codex/agents/context-search-agent.toml

Lines changed: 8 additions & 17 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,10 @@
1-
---
2-
name: context-search-agent
3-
description: |
4-
Intelligent context collector for development tasks. Executes multi-layer file discovery, dependency analysis, and generates standardized context packages with conflict risk assessment.
5-
6-
Examples:
7-
- Context: Task with session metadata
8-
user: "Gather context for implementing user authentication"
9-
assistant: "I'll analyze project structure, discover relevant files, and generate context package"
10-
commentary: Execute autonomous discovery with 3-source strategy
11-
12-
- Context: External research needed
13-
user: "Collect context for Stripe payment integration"
14-
assistant: "I'll search codebase, use Exa for API patterns, and build dependency graph"
15-
commentary: Combine local search with external research
16-
color: green
17-
---
1+
name = "context_search_agent"
2+
description = "Intelligent context collector for development tasks. Executes multi-layer file discovery, dependency analysis, and generates standardized context packages with conflict risk assessment."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "read-only"
6+
7+
developer_instructions = """
188
199
You are a context discovery specialist focused on gathering relevant project information for development tasks. Execute multi-layer discovery autonomously to build comprehensive context packages.
2010
@@ -587,3 +577,4 @@ Output: .workflow/session/{session}/.process/context-package.json
587577
### Windows Path Format Guidelines
588578
- **Quick Ref**: `C:\Users` → MCP: `C:\\Users` | Bash: `/c/Users` or `C:/Users`
589579
- **Context Package**: Use project-relative paths (e.g., `src/auth/service.ts`)
580+
"""

.codex/agents/debug-explore-agent.md renamed to .codex/agents/debug-explore-agent.toml

Lines changed: 8 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,10 @@
1-
---
2-
name: debug-explore-agent
3-
description: |
4-
Hypothesis-driven debugging agent with NDJSON logging, CLI-assisted analysis, and iterative verification.
5-
Orchestrates 5-phase workflow: Bug Analysis → Hypothesis Generation → Instrumentation → Log Analysis → Fix Verification
6-
color: orange
7-
---
1+
name = "debug_explore_agent"
2+
description = "Hypothesis-driven debugging agent with NDJSON logging, CLI-assisted analysis, and iterative verification."
3+
model = "gpt-5.4"
4+
model_reasoning_effort = "high"
5+
sandbox_mode = "workspace-write"
6+
7+
developer_instructions = """
88
99
You are an intelligent debugging specialist that autonomously diagnoses bugs through evidence-based hypothesis testing and CLI-assisted analysis.
1010
@@ -434,3 +434,4 @@ ${nextSteps}
434434
- Timeout: Analysis 20min | Fix implementation 40min
435435
436436
---
437+
"""

0 commit comments

Comments
 (0)