skill-builder

# /skill-builder — Scaffold or absorb a new SKILL.md

Materializes a new skill against the unified template at `references/skill-template.md` (extracted from anthropics/financial-services). Runs `skill-auditor` on the new skill as a self-check before declaring success.

> **If unsure whether the work should be a skill, a Workflow, or an NTM swarm, run `/automation-shape-routing` first** — it is the front door that decides the shape and hands off to the right builder.

## ⚠️ Critical Constraints

- **Template is canonical.** All four modes produce SKILL.md files conforming to `references/skill-template.md`. Do not invent ad-hoc structures. **Why:** `skill-auditor` validates against this template; drift creates auditor false-fails.
- **Self-audit is mandatory.** After every successful build, the build script invokes `/skill-auditor` against the new skill directory. A FAIL verdict aborts the build. **Why:** PR-002 (external validation gate) — the builder must not declare its own work complete.
- **Codex parity is day-1, not later.** `from-scratch`, `from-template`, and `absorb-external` modes must produce both `skills/<name>/SKILL.md` AND `skills-codex/<name>/SKILL.md` + `skills-codex/<name>/prompt.md`. **Why:** finding `2026-05-03-codex-skill-shape-is-dual-file` — codex SKILL.md uses slim frontmatter (no `skill_api_version`); prompt.md is mandatory; `audit-codex-parity.sh` is a content scanner that won't catch frontmatter drift.
- **250-line ceiling on new SKILL.md.** Use `references/` for overflow. **Why:** finding `f-2026-05-01-025` — every Skill() invocation reloads 5-15KB; multi-lifecycle sessions compound to 150-200KB+ pure scaffolding.
- **Clean-room factory inputs only.** When using lessons learned from external corpora, read [references/agentops-skill-factory.md](references/agentops-skill-factory.md) and use only AgentOps-owned summaries, scripts, and rubrics. **Why:** productization must improve structure without copying protected third-party skill content.
- **Real gate means exit code.** Validate with `heal-skill --check --strict <skill-dir>` and `skill-auditor`; never infer green from grep/regex output. **Why:** regex presence checks created false-greens during the 2026-06 scale build.
- **One skill directory = one writer.** Bulk builds fan out only when each worker owns a distinct new `skills/<name>/` plus `skills-codex/<name>/`; edits to existing skill dirs run in a later serial wave. **Why:** concurrent writers deleted untracked work and flipped HEAD mid-task.
- **Trust repository state, not subagent reports.** Before declaring success, inspect `git status`, generated hashes, final files, and gate exit codes. **Why:** sandbox-overlay and stale self-reports can claim work that never persisted.
- **Clean-room includes names.** Do not reuse exact third-party skill names; mint AgentOps-owned names before source skills, Codex mirrors, or wrappers are keyed. **Why:** provenance/IP safety applies to labels as well as prose and scripts.
- **Do not use the Workflow tool as the skill factory.** For scale authoring, use deterministic wave scripts or NTM/Agent Mail lanes with one worker per skill. **Why:** skill creation needs file ownership and durable git evidence, not opaque background self-reporting.

## Modes

| Mode | Status | Description |
|------|--------|-------------|
| `from-scratch` | stable | Interactive scaffold from canonical template. Produces full skill skeleton + scripts/validate.sh + codex parity. |
| `from-template` | stable | `--like <existing-skill>` copies structure from a sibling skill, swaps domain-specific sections. |
| `absorb-external` | stable | Reads external SKILL.md (e.g., from `~/dev/financial-services/<some-dir>/<skill>/SKILL.md`), wraps in AgentOps frontmatter, invokes `/converter` for codex parity. |
| `from-pattern` | **alpha (passthrough)** | Delegates to `ao flywheel close-loop`. Outputs land at `.agents/knowledge/promoted/` per flywheel rules — they are NOT yet shaped as SKILL.md drafts. v2 will add skill-specific synthesis. Use `from-scratch` or `absorb-external` for SKILL.md output today. |

## Workflow

### Phase 1: Mode dispatch

`scripts/build.sh` reads `$1` and routes:

```bash
build.sh from-scratch <new-skill-name>          # → init.sh --interactive
build.sh from-template <new-skill-name> --like council
build.sh absorb-external <new-skill-name> --from /path/to/SKILL.md
build.sh from-pattern                            # → ao flywheel close-loop
```

**Checkpoint:** Confirm with user the new skill's `metadata.tier` and `metadata.dependencies` before generation.

### Phase 2: Materialize from template

`scripts/init.sh` reads `references/skill-template.md` (the canonical template section) and renders a SKILL.md skeleton with frontmatter pre-filled. For `from-template`, structure is copied from the source skill; section bodies are blanked and replaced with template stubs.

For `absorb-external`, the external SKILL.md's content (Constraints / Workflow / Output / Quality sections) is preserved verbatim where possible; AgentOps' structured frontmatter is added on top; the external description is reformatted to satisfy `description-has-triggers`.

**Checkpoint:** `heal-skill --check --strict skills/<new-name>` exits 0.

### Phase 3: Codex parity

`scripts/init.sh` invokes `/converter skills/<new-name> codex` to produce `skills-codex/<new-name>/{SKILL.md,prompt.md}`. Then trims `skill_api_version` from the codex SKILL.md (converter may preserve it). Asserts `prompt.md` exists.

**Checkpoint:** `bash scripts/audit-codex-parity.sh` returns clean AND `grep -q "^skill_api_version:" skills-codex/<name>/SKILL.md` returns nothing.

### Phase 4: Self-audit

The build script tail invokes `/skill-auditor` on `skills/<new-name>`. WARN is acceptable for v1 skills (e.g., `experimental` stability). FAIL aborts.

**Checkpoint:** `audit_pass=true` in build report.

### Phase 5: Factory score overlay

For AgentOps skill upgrades, use the productization score as a p