Agent Skill for shell script modules under a repository's .scripts/ directory — the api/ + lib/ split, the --index.sh bootstrap chain, Taskfile integration, and the shared loggers and base foundation modules.

Layout

development-scripts/
├── SKILL.md                          # Entry point — routing, anatomy, naming, runtime flow
├── references/
│   ├── module-anatomy.md             # Reader, validator, index files, implementations, wrappers
│   ├── shared-modules.md             # loggers and base, and the logging bootstrap rule
│   ├── taskfile.md                   # Module and root Taskfile integration
│   ├── workflows.md                  # Create, change, validate
│   ├── module-patterns.md            # Module type and dependency decision guide
│   └── environment-model.md          # Layered .env model, samples, shipping into a target repo
└── assets/
    ├── module-example-v1/            # Scaffold to copy for a new module
    ├── loggers/                      # Working reference implementation
    └── base/                         # Working reference implementation

The model

<target-repository>/
  Taskfile.yml
  .scripts/
    <module>/
      api/                 # public command wrappers — what users run
      lib/                 # implementation functions — what the code calls
      Taskfile.yml         # only when exposed through task

Commands run from the repository root:

./.scripts/<module>/api/<command>.sh --dry-run
task <module-alias>:<task-name> -- --dry-run

Every source path is repository-root-relative, and the bootstrap chain — reader, then validator, then API index — guarantees an implementation function sees a fully-prepared environment.

Deploy

cd ../skill-manager/scripts
task deploy -- --skill-dir="$(cd ../../development-scripts && pwd)"

Built with the skill-manager skill, following the agentskills.io specification.