add-framework-internal-layerlisted

# Internal Layer Mechanics <!-- uses: - skill: building-commands - skill: add-framework-development - mention: add-build-ledger - mention: /add-framework--done --> Loaded on the first `[internal]` F-block of a build. These artefacts are the development tooling that builds and maintains the framework. **They are NOT distributed to users.** WHAT gets recorded about execution is `add-build-ledger`'s job; the product layer has its own skill. ## When to Use - An F-block naming `.claude/`, `scripts/`, `CLAUDE.md`, `.gitignore`, or a repo-root file. ## When NOT to Use - An F-block naming `framwork/.codeadd/`, `framwork/provider-map.json` or `cli/`. --- ## What the Internal Layer Is | Type | Path | |------|------| | Commands | `.claude/commands/*.md`, namespace `add-framework--*` | | Skills | `.claude/skills/<name>/SKILL.md`, subdocs in `references/` | | Agents | `.claude/agents/*.md` | | Support scripts | `scripts/*.js` | | Structural map | `CLAUDE.md` | ``` ⛔ INTERNAL ARTEFACTS ARE NOT REGISTERED: ⛔ DO NOT: Add an internal command, skill or agent to framwork/provider-map.json ⛔ DO NOT: Copy an internal artefact into framwork/.codeadd/ ✅ DO: Create it directly at its .claude/ path — build.js never distributes it ``` **The internal layer has no provider mirror.** One file per artefact, no adapter to keep in step. `add-framework-development` carries the artefact anatomies, the agent frontmatter fields and the `<!-- uses: -->` syntax. Read it when creating a new inter