---
name: lythoskill-deck
version: 0.19.1
description: |
  Declarative skill deck governance. `deck link` reconciles the working set
  to match skill-deck.toml — deny-by-default removes undeclared skills.
  DEFAULT patterns: PHASE SWITCH via separate deck files (deck link --deck
  phase<N>.toml), SEED bootstrap (minimal deck → agent self-expands via
  deck add + curator discovery). Always restores parent deck. Zero state
  pollution between phases. INNATE FIRST: after compaction, session reload,
  or skill refresh — read every innate skill's full SKILL.md before any
  tool skill. Innate = eagerly loaded, always full context. Tool = lazy,
  read only on trigger. CRITICAL when writing a deck for others: verify
  every locator path against the REAL repo structure before publishing.
when_to_use: |
  Skill conflicts, too many skills, clean up or organize skills,
  skill-deck.toml, sync working set, initialize deck, manage skill deck,
  phase switch, seed bootstrap, fork to localhost, multi-file deck,
  deck add, deck link, deck refresh, deck validate, deck remove,
  silent blend, what are these symlinks in .claude/skills/.
  ALSO trigger when user says "switch deck", "add skill", "remove skill",
  "clean up skills", "organize my deck", "create phase deck".
allowed-tools:
  - Bash(bunx @lythos/skill-deck@0.19.1 *)
# ── deck governance metadata (consumed by lythoskill tooling, not by agent platforms) ──
deck_niche: meta.governance.deck
deck_dependencies:
  runtime: [bash]
  optional: [bun]
deck_managed_dirs:
  - .claude/skills/
  - skill-deck.lock
---

# lythoskill-deck: Declarative Skill Deck Governance
> What matters is not how many skills you have, but which ones the agent sees at the same time.

## Core Model

`skill-deck.toml` declares desired state. `deck link` makes the working set (default `.claude/skills/`) match — creates symlinks for declared skills, removes everything else. Deny-by-default: undeclared skills do not exist in the agent's view.

```
Cold Pool (~/.agents/skill-repos/)  →  deck add  →  skill-deck.toml  →  deck link  →  .claude/skills/
     (all downloaded skills)              (declare)      (desired state)    (reconcile)    (working set — what agent sees)
```

## Multi-File Phase Decks (BEST PRACTICE)

When a task spans different skill sets, use **separate deck files per phase** instead of editing a single toml:

```bash
# Phase 1 — brand design
deck link --deck phase1-brand.toml --cold-pool ~/.agents/skill-repos

# Phase 2 — content creation (atomic switch, Phase 1 skills gone)
deck link --deck phase2-content.toml --cold-pool ~/.agents/skill-repos

# Done — restore parent deck
deck link --deck ./skill-deck.toml
```

Each phase deck is independently auditable. Never add/remove entries in-place when switching contexts — a wrong edit silently breaks the previous phase's composition. The reconciler handles the transition: old symlinks removed, new ones created, no state leaks.

## Seed Bootstrap

Start with a minimal deck (only lythoskill-deck as innate). The agent reads deck SKILL.md → learns schema → uses curator `name LIKE '%keyword%'` to discover skills → `deck add` + `deck link` → self-expands. Governance skill is the only irreducible dependency.

```toml
[deck]
max_cards = 8
cold_pool = "~/.agents/skill-repos"
working_set = ".claude/skills"

[innate.skills.lythoskill-deck]
path = "github.com/lythos-labs/lythoskill/skills/lythoskill-deck"
```

## Commands

```bash
# ── Core (use these every time) ──

# Reconcile working set to match declaration
bunx @lythos/skill-deck@0.19.1 link
bunx @lythos/skill-deck@0.19.1 link --deck ./phase1-brand.toml

# Add skill from cold pool or download URL
bunx @lythos/skill-deck@0.19.1 add github.com/owner/repo/skill-name
bunx @lythos/skill-deck@0.19.1 add github.com/owner/repo/skill-name --alias my-skill --type tool

# ── Maintenance ──

bunx @lythos/skill-deck@0.19.1 refresh              # plan-only scan for updates
bunx @lythos/skill-deck@0.19.1 refresh --exec       # actually git pull (self-heals dirty cold pool; non-zero exit + trailing ⚠️ on failure)
bunx @lythos/skill-deck@0.19.1 remove <alias>       # remove from deck + working set
bunx @lythos/skill-deck@0.19.1 validate                   # check current deck
bunx @lythos/skill-deck@0.19.1 validate --deck <path>     # check specific deck
bunx @lythos/skill-deck@0.19.1 validate --remote          # probe locators against GitHub
# Cold-pool-missing skills warn (not error) — they may exist on GitHub.
# HATEOAS output: each warning suggests curator add <locator> as next step.

# ── Advanced ──

bunx @lythos/skill-deck@0.19.1 to-symlink <alias>   # snapshot → symlink
bunx @lythos/skill-deck@0.19.1 to-snapshot <alias>  # symlink → snapshot (pin HEAD)
bunx @lythos/skill-deck@0.19.1 reconcile            # drift report vs cold pool
bunx @lythos/skill-deck@0.19.1 reconcile --apply    # converge
bunx @lythos/skill-deck@0.19.1 migrate-schema       # old string-array → alias-as-key
```

`link` is a reconciler: undeclared symlinks → removed; broken symlinks → recreated; non-symlink entities → backed up then removed; missing declared skills → linked from cold pool. It also prints best-effort ⚠️ warnings when a cold-pool repo is behind origin, dirty, or on a non-default branch — drift is surfaced at the step every boot already runs. The health probe never blocks and swallows its own errors by design; set `LYTHOS_DEBUG=1` to print a caught probe error (🔍 whisper) when you suspect the probe itself is broken.

`refresh` defaults to **plan-only** (no git pull). Use `--exec` to apply, or let an agent read the plan and execute per target — the agent can probe remotes, switch mirrors, handle divergence.

> `deck update` is deprecated. Use `refresh`.

## Format Detection (agent duty)

Before running any `deck` command, check the toml format. If you see `skills = [...]` (string array), it's the deprecated format. Ask before migrating:

```
⚠️  This deck uses the deprecated string-array format. Run migrate?
   bunx @lythos/skill-deck@0.19.1 migrate-schema
   bunx @lythos/skill-deck@0.19.1 migrate-schema --dry-run  # preview first
```

Do NOT silently migrate.

## Key Concepts

| Concept | One-liner |
|---------|-----------|
| Cold Pool | All downloaded skills (`~/.agents/skill-repos/`). Agent cannot see here. |
| skill-deck.toml | Declares desired state. Alias-as-key dict format. |
| deck link | Reconciler. Makes the working set match the declaration. |
| Working Set | Symlinks only. What the agent actually scans. Default: `.claude/skills/` (configurable per platform). |
| skill-deck.lock | Machine-generated snapshot: resolved paths, hashes, constraints. |

## lock vs state — what to commit and why
The deck writes two machine files with opposite git policies
(ADR-20260616000939948 — born from a real incident: a ZK agent saw the old
all-in-one lock dirty after a routine link and asked whether to commit).

| File | Content-derived? | git | Why |
|------|-----------------|-----|-----|
| `skill-deck.toml` | hand-edited desired state | **commit** | The declaration IS the reproducibility contract (`go.mod` role). |
| `skill-deck.lock` | yes — relative paths, FQ locators, content hashes only; no timestamps, no absolute paths | **commit** (root lock) | `go.sum`/`bun.lockb` role: pins "which content was linked", so any machine/agent resolves identical bytes, probe detects drift, agent B inherits agent A's resolution. |
| `skill-deck.state` | no — `generated_at`/`linked_at` timestamps, absolute paths, symlink dest/mode | **ignore** | Changes on EVERY link and encodes this machine's filesystem. Regenerate anytime via `deck link`. Never read it for "what version is locked" — that is the lock's job. |

Also ignore: working sets (`.claude/`, `.agents/skills/` — symlinks are
per-machine) and non-root experiment lockfiles (`.agents/skill-deck.lock`,
`showcase/**/skill-deck.lock`). This repo's `.gitignore` is the reference
implementation.

**Why the lock belongs in git**: a hash-only lock diff is the audit trail
of "upstream skill content moved" — the alternative is silent drift. The
split exists precisely so the lock behaves like a lockfile (changes only
on content change) while link's operational churn lands in the ignored
state file. `deck link` prints a drift hint when the lock changes — follow
it; it stays silent when nothing moved.

**Wrap-up ordering (real case, 2026-08-31 release)**: the session's final
`deck link` ran at 17:46:55 — after the scribe's git-status snapshot, 80 s
before the daily commit. Nobody re-checked `git status`; the lock sat
dirty for a week and cost the next onboarding a judgment cycle. Rule: any
link/refresh is followed by a fresh `git status` as the LAST operation
before the final commit.

## Constraints

- **deny-by-default** — undeclared skills are physically absent from working set
- **max_cards** — exceeding the budget causes link to refuse
- **link backs up real directories** — non-symlink entries archived to `.claude/skills.bak.*.tar.gz`
- **transient expires** — past-due transients trigger warnings
- **managed_dirs overlap** — two skills claiming same directory triggers warning
- **Never manually create subdirectories in the working set directory** — use `deck link`
- **deck does not run post-install steps** — API keys, env vars are the skill's own responsibility

## Locators — Always Fully-Qualified

| Style | Example | Reliability |
|-------|---------|-------------|
| Fully-qualified | `github.com/lythos-labs/lythoskill/skills/lythoskill-deck` | Reliable — deterministic path |
| Bare name | `lythoskill-deck` | Fragile — non-deterministic readdir order |

The deck does NOT auto-insert a `skills/` prefix. Verify repo structure before writing locators.

## Gotchas

**deck link uses lstatSync, not existsSync** — `existsSync` returns false for broken symlinks, causing EEXIST errors.

**SKILL.md type field**: Only `standard` or `flow` are valid. `innate`/`tool`/`transient` are deck toml section names, not SKILL.md types. `deck_skill_type` (custom field) for `fork`/`transient` declarations.

**Network restrictions**: The pre-clone connectivity probe is advisory — `deck add` always attempts the clone and lets git's own error be authoritative (probe false negatives on git smart-HTTP mirrors were a real incident). If a clone fails with network errors, use one of these env vars:

| Variable | When to use | Example |
|----------|-------------|---------|
| `LYTHOS_GH_MIRROR` | You have a GitHub mirror/proxy | `export LYTHOS_GH_MIRROR="https://mirror.example.com"` |
| `LYTHOS_SOCKS_PROXY` | You have a SOCKS5 proxy | `export LYTHOS_SOCKS_PROXY="127.0.0.1:1080"` |
| `LYTHOS_GIT_PROTOCOL` | You need SSH instead of HTTPS | `export LYTHOS_GIT_PROTOCOL="ssh"` |

```bash
export LYTHOS_SOCKS_PROXY="127.0.0.1:1080"
bunx @lythos/skill-deck@0.19.1 add github.com/owner/repo/skill
```

**Innate priority**: After compaction, read `innate` skills' full SKILL.md first. `tool` skills are lazy — read on trigger. Agent-side convention.

**Never guess locators** — web-search the repo structure before writing paths for unfamiliar repos.

**skill-deck.lock goes dirty by design** — `link`/`refresh` recompute the lock's content hashes from linked skill content, which lives OUTSIDE the repo (working-set symlinks → cold pool / upstream). The lock is the only in-repo artifact that moves, so `git status` can show `M skill-deck.lock` while no repo file changed. That is derived-state drift, not repo drift — don't burn a judgment cycle on it. Triage via `git diff skill-deck.lock`: hash-only lines = external content moved, safe to commit as housekeeping; path/entry/structure changes = real working-set change, review before committing. Never hand-edit the lock — regenerate via `deck link` (reverting `skill-deck.toml` does not revert derived state; re-run `link`).

## Deck as Orchestrator — Task → Deck Mapping

lythos has no standalone orchestrator. **The deck IS the orchestrator entry point.** You (the agent) select the right pre-built deck for the user's task, link it in an isolated environment, and execute. The three-layer distribution: light orchestration in combo prompt, medium in SKILL.md, heavy mechanical in CLI.

**When you encounter a deck with `[combo.<name>]` sections, READ the combo prompts.** They are the orchestration playbook — step-by-step instructions for multi-skill pipelines. The combo section is not optional metadata; it IS the orchestrator. Skipping it means executing skills in isolation without the coordination logic.

### Intent → Deck mapping

| User says | Deck to use | Command |
|-----------|------------|---------|
| 调研 / 研究 / 查一下 ("research" / "look into it") / research | `deep-research.toml` | `arena single --deck examples/decks/deep-research.toml --brief "..."` |
| 扫一下 / 审计 / 找问题 ("sweep" / "audit" / "find problems") / audit / sweep | `qa-sweep.toml` | `arena single --deck examples/decks/qa-sweep.toml --brief "..."` |
| 设计 / 架构 / 画图 ("design" / "architecture" / "draw diagrams") / architecture | `architecture-explainer.toml` | `arena single --deck examples/decks/architecture-explainer.toml --brief "..."` |
| 写文档 / 科普 ("write docs" / "explain to a general audience") / docs | `documents.toml` | `arena single --deck examples/decks/documents.toml --brief "..."` |
| 治理 / task / epic ("governance") / governance | `governance.toml` | `arena single --deck examples/decks/governance.toml --brief "..."` |
| 体验 / 测试 UX ("try it out" / "UX-test") / scout | `scout.toml` | `arena single --deck examples/decks/scout.toml --brief "..."` |
| 改bug / 诊断 ("fix bugs" / "diagnose") / diagnose | Use `diagnose` skill directly | — |
| 写测试 / TDD ("write tests") | Use `tdd` skill directly | — |
| 发布 / release / ship ("release") | Use `lythoskill-red-green-release` skill directly | — |

**Rule: deck first, don't ask.** User says "调研" ("research it") → reach for
`deep-research.toml`. User says "扫一下" ("sweep it") → `qa-sweep.toml`. Do NOT
ask "should I use X or do it manually?" — just dispatch. The subagent works in parallel.

Full catalogue: [`examples/decks/INDEX.md`](./examples/decks/INDEX.md).

### Isolation discipline

**Experiments run in `/tmp`, never in committed directories.** Arena's `prepare-workdir` creates an isolated temp dir, links the deck there, and archives only outputs. No pollution of the project working set.

```bash
# Golden rule: never run experiments in the project root
arena single --deck examples/decks/deep-research.toml --brief "..."  # auto-isolates in /tmp
```

### Orchestration weight — where does the logic go?

| Your orchestration logic is… | Put it in… | Example |
|------------------------------|------------|---------|
| Light + conditional ("if X then Y") | `[combo.<name>] prompt` in skill-deck.toml | "If git status shows conflicts, call merge-conflict" |
| Medium + reusable across projects | Standalone SKILL.md in cold pool | A complex workflow with branching logic |
| Heavy + mechanical (filesystem, backup) | CLI npm package | `deck link`, `arena single` |

**The agent is the orchestrator. Tools execute.** You reason about conditions, pass state between calls, handle errors. Combo prompt gives you the playbook; CLI gives you the mechanical layer.

## Supporting References

| When you need to… | Read |
|--------------------|------|
| Understand cold pool → deck → working set pipeline | [references/architecture.md](./references/architecture.md) |
| Look up a lythoskill term | [references/glossary.md](./references/glossary.md) |
| Write or edit skill-deck.toml | [references/toml-format.md](./references/toml-format.md) |
| Set up a cold pool | [references/cold-pool-setup.md](./references/cold-pool-setup.md) |
| Build or organize a deck | [references/deck-building.md](./references/deck-building.md) |
| Run arena benchmarks with deck isolation | [references/arena-integration.md](./references/arena-integration.md) |
| Understand SKILL.md types or skill thickness | [references/skill-types.md](./references/skill-types.md) |
| Add custom frontmatter fields | [references/custom-fields.md](./references/custom-fields.md) |