Group G — markdown tech-debt cleanup (post-audit 2026-05-02).
- 36 SKILL.md files: added "## When to use" section. Was missing across the
catalog; orchestrator routing by keyword could not auto-dispatch.
- 20 code-implementer agent .md files: added Output Footer block prescribing
RULE 0.16 STATUS-TRUTH MARKER schema in agent's final report. Previously only
code-implementer-rust.md had it; other 27 language/role variants were silent
about the marker, breaking RULE 0.16 §3 status-truth aggregation for non-Rust
batches.
- skills/site-create/: added phase-5-preview.md and phase-6-deploy.md skeleton
files. SKILL.md table-of-contents referenced 7 phases; only 5 existed on disk.
- skills/{ai-animation,rag-pipeline}/skill.md: added migration banner comment
noting they should be SKILL.md (canonical filename). Case-rename via git is a
separate orchestrator task (macOS APFS is case-insensitive; Linux deploy needs
explicit rename).
- 3 deprecated skills (site-builder, competitor-analysis, design-inspiration):
added concrete removed-after dates (was vague "before v2").
- docs/CONVERGENCE-PLAN.md:129: TBD on _blocks/evidence-grading.md duplicate
resolved (file exists, not duplicated).
- docs/DNA-INDEX.md: count edits made then overwritten by auto-encyclopedia-refresh
hook during agent run. The .kei-registry-ignore files in test fixtures (Group F)
are the structural fix; kei-registry walker implementation is the follow-up.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
4 KiB
4 KiB
| name | description | category |
|---|---|---|
| pet-init | Create a personal AI pet persona via interactive wizard. No TOML editing required. | pet |
Pet Init — Interactive Persona Wizard (index)
When to use
- Creating a personal AI pet persona via interactive wizard — no TOML editing required.
- Setting up a
pet.tomlmanifest forkei-petwith custom name, voice, tone, and edge-case settings. - First-time persona creation or resetting an existing persona to new preferences.
You are helping a non-developer create their personal AI pet persona. The
output is a valid pet.toml manifest conforming to the kei-pet schema
(see _primitives/_rust/kei-pet/examples/minimal.toml). The user NEVER
edits TOML by hand — every field is gathered through AskUserQuestion
batches or short free-text prompts.
This SKILL.md is the INDEX. Each phase lives in its own file and runs in
strict order. Never skip or re-order phases.
Pipeline overview (4 phases)
| Phase | File | Purpose | AskUserQuestion |
|---|---|---|---|
| 1 | phase-1-identity.md | Pet name, user name, addressing style, languages | 1 batch (2 questions) |
| 2 | phase-2-voice.md | Tone, humor style, humor frequency | 1 batch (4 questions) |
| 3 | phase-3-edge.md | Directness, initiative, profanity, forbidden topics | 1 batch (3 questions) + 1 free-text |
| 4 | phase-4-emit.md | Compose TOML, keygen if needed, write file, summary | 0 |
Exit: ~/.claude/pet/<user_id>.toml written, summary displayed, next-step
suggestion shown.
Variables the pipeline produces
| Name | Set in | Shape |
|---|---|---|
PET_NAME |
Phase 1 | string, 1-30 chars |
USER_NAME |
Phase 1 | string |
ADDRESSING |
Phase 1 | by-name / formal / casual |
LANGUAGES |
Phase 1 | array of ISO codes |
TONE_PRIMARY |
Phase 2 | warm / neutral / formal / playful |
TONE_SECONDARY |
Phase 2 | array, 0-2 entries |
HUMOR_STYLE |
Phase 2 | none / dry / witty / silly |
HUMOR_FREQUENCY |
Phase 2 | rare / occasional / frequent |
DIRECTNESS |
Phase 3 | gentle / balanced / direct / blunt |
INITIATIVE |
Phase 3 | wait / nudge / proactive |
PROFANITY |
Phase 3 | never / rare / contextual |
FORBIDDEN_TOPICS |
Phase 3 | array of strings, may be empty |
USER_ID |
Phase 4 | Ed25519 short id (from kei-pet keygen) |
Rules (apply throughout)
- No manual TOML. The user never sees or edits raw TOML until after
Phase 4 emits the file. Any correction = re-run
/pet-init. - RULE 0.4 (NO HALLUCINATION). Never invent defaults silently. Every field is either asked or explicitly defaulted in the phase file.
- RULE 0.8 (SECRETS). The Ed25519 secret key (created by
kei-pet keygen) is written by the primitive into its own keystore — this skill never reads or displays secret-key material. Only the publicuser_idshort-hash is surfaced to the user. - NO DOWNGRADE. If Phase 4 cannot write the file (permission error, disk full, keygen failure), return 2-3 constructive paths — never "can't be done".
- Constructor Pattern. Each phase file is a single cube ≤200 LOC. This index stays ≤200 LOC.
- Surgical Changes. The only file written by this skill is
~/.claude/pet/<user_id>.toml. No other artefacts.
Exit report (emit after Phase 4)
=== PET-INIT REPORT ===
Pet name: <PET_NAME>
User: <USER_NAME>
File: ~/.claude/pet/<USER_ID>.toml
Size: <bytes>
Keygen: <reused existing | newly created>
Next: /pet-chat or kei-pet render --pet ~/.claude/pet/<USER_ID>.toml
References
- phase-1-identity.md
- phase-2-voice.md
- phase-3-edge.md
- phase-4-emit.md
_primitives/_rust/kei-pet/examples/minimal.toml— schema reference_primitives/_rust/kei-pet/examples/full.toml— optional-section reference