talont-org/planner-analyze

planner-analyze

Detect language, framework, test infrastructure, project structure, and existing patterns in the target project. Produces analysis.json used by subsequent planner skills.

When to Use

• Invoked by the planner recipe as the first analysis step
• User says "analyze project structure" in a planning context

Arguments

• $1 — Absolute path to the run-scoped planner directory (e.g., {{AUTOSKILLIT_TEMP}}/planner/run-YYYYMMDD-HHMMSS). Created by the init step.

Critical Constraints

NEVER:

• Modify any target project files
• Write analysis.json outside $1/
• Run subagents in the background (run_in_background: true is prohibited)

ALWAYS:

• Use Explore subagents for all file reads
• Spawn all 4 subagents in parallel
• Write valid JSON to analysis.json

Workflow

Step 1: Launch 4 parallel Explore subagents

Spawn all four concurrently with model: "sonnet":

1. Languages & Frameworks — Identify primary language, framework, build system. Look for: pyproject.toml, package.json, Cargo.toml, go.mod, pom.xml, build.gradle, import statements, dependency files.

2. Test Infrastructure — Identify test runner, coverage tools, test directory layout. Look for: pytest.ini, jest.config.*, go test, cargo test, test file naming patterns.

3. Architecture Patterns — Identify architecture style (layered, hexagonal, monolithic, microservices, etc.) and count modules. Look for: directory depth, import graphs, layer naming, package boundaries.

4. Existing Conventions — Identify naming conventions, code patterns, and risk areas. Look for: consistent naming in identifiers, repeated structural patterns, areas with high coupling or missing tests.

Step 2: Synthesize results

Merge all four agent outputs into a single analysis.json document matching the output schema.

Step 3: Write output

Write to $1/analysis.json. The directory was created by the init step.

Output Schema

{
  "language": "python",
  "framework": "fastapi",
  "build_system": "uv",
  "test_runner": "pytest",
  "architecture_style": "layered",
  "module_count": 42,
  "key_patterns": ["dependency injection", "protocol-based contracts"],
  "conventions": ["snake_case identifiers", "private prefix _"],
  "risks": ["high coupling in server layer", "no tests for migration engine"]
}

All fields are required. Use null for fields that cannot be determined. Arrays may be empty but must be present.