@hegemonart/get-design-done 1.57.2 → 1.57.3

package/dist/claude-code/.claude/skills/apply-reflections/SKILL.md

@@ -1,109 +0,0 @@
----
-name: gdd-apply-reflections
-description: "Review and selectively apply proposals from .design/reflections/<cycle-slug>.md. Diffs each proposal, prompts user to accept/skip/edit, then writes changes."
-argument-hint: "[--cycle <slug>] [--filter <FRONTMATTER|REFERENCE|BUDGET|QUESTION|GLOBAL-SKILL>] [--dry-run]"
-tools: Read, Write, Edit, Bash, Glob
----
-
-# /gdd:apply-reflections
-
-Interactive proposal review loop. Reads `.design/reflections/<cycle-slug>.md`, walks each numbered proposal, and applies accepted ones to the appropriate target file. Nothing is applied without explicit user confirmation.
-
-## Steps
-
-### 1. Resolve reflections file
-
-- If `--cycle <slug>` given: load `.design/reflections/<slug>.md`
-- Else: glob `.design/reflections/*.md`, sort by modified time descending, load the most recent
-- If no file found: error "No reflections found. Run `/gdd:reflect` first."
-- Print: "Reviewing reflections: <filename>"
-
-### 2. Parse proposals
-
-Scan file for lines matching `### Proposal N — [TYPE] ...`. Extract each proposal block (Why / Change / Risk).
-
-If `--filter <TYPE>` given: skip proposals whose type tag doesn't match.
-
-Print: "Found N proposals (N after filter)."
-
-### 3. Review loop
-
-For each proposal (in order):
-
-Print the full proposal block:
-```
-─────────────────────────────────────────
-Proposal N/TOTAL — [TYPE] Title
-Risk: low|medium
-
-Why: ...
-Change: ...
-─────────────────────────────────────────
-(a) apply   (s) skip   (e) edit   (q) quit
-```
-
-If `--dry-run`: print `[dry-run — would prompt here]` and continue to next proposal without prompting.
-
-Based on user choice:
-- **a** - apply (see Apply Logic below)
-- **s** - mark proposal as `**Reviewed: skipped**` in the reflections file; continue
-- **e** - show the Change text, ask user to provide edited version, then apply the edited version
-- **q** - stop processing; print "Stopped at proposal N. Resume with `/gdd:apply-reflections --cycle <slug>`."
-
-### 4. Apply Logic by Proposal Type
-
-After the user chooses `a` (apply) or `e` (edit-then-apply), branch on the proposal's bracketed type tag and follow the per-type apply procedure in `./apply-reflections-procedure.md` - one numbered procedure each for `[FRONTMATTER]`, `[REFERENCE]`, `[BUDGET]`, `[QUESTION]`, `[GLOBAL-SKILL]`. All branches end with `**Applied**: <date>` appended to the proposal block in the reflections file.
-
-### 5. Summary
-
-After all proposals processed (or `q`):
-```
-─────────────────────────────────────────
-Apply-reflections complete
-  Applied:  N
-  Skipped:  N
-  Remaining: N (run again to continue)
-─────────────────────────────────────────
-```
-
-## [INCUBATOR]
-
-Incubator drafts authored by `scripts/lib/incubator-author.cjs` (Phase 29-04) appear as a distinct proposal class. For each draft under `.design/reflections/incubator/<slug>/`, use `scripts/lib/apply-reflections/incubator-proposals.cjs`:
-
-1. `discoverIncubatorDrafts()` → list pending drafts.
-2. `renderProposal(draft)` → show full body + diff + origin signals.
-3. User chooses **accept** | **reject** | **defer** | **edit**.
-4. **accept** - scope-guard runs FIRST (`validateScope` from `scripts/validate-incubator-scope.cjs`); `applyAccept` then promotes draft → `agents/<slug>.md` or `skills/<slug>/SKILL.md` and appends a registry entry. Single-step per D-04.
-5. **reject** - `applyReject` removes the incubator subdir.
-6. **defer** - no-op; draft re-surfaces next run.
-7. **edit** - `applyEdit` opens `$EDITOR`; re-prompt user on close.
-
-**Stage-1 gate.** At session start, call `checkStage1Gate()`. If `thresholdMet && !optInRecorded`, display the opt-in prompt once. NEVER auto-flip per D-01 - recording opt-in requires explicit user confirmation via `recordOptIn()`. Full procedure: `./apply-reflections-procedure.md` §[INCUBATOR].
-
-## [KFM-CANDIDATE]
-
-KFM-catalogue proposals authored by `scripts/lib/reflector-kfm-proposer.cjs` (Phase 30.5-03 D-05) appear as a 6th proposal class. Drafts at `.design/reflections/incubator/kfm-<slug>/CATALOGUE-ENTRY.md`; pre-filled 11-field schema with `TODO:` placeholders for `pattern` + `fix`. Two upstream signals share the surface (D-06): `capability_gap` clusters (≥3, no existing match) + `kfm-candidate` events (whitelist-matched articles, 1-shot). User chooses **accept** | **reject** | **defer** | **edit**. `applyAccept` appends to `reference/known-failure-modes.md` + `reference/registry.json` (`origin: incubator-kfm`); `applyReject` removes the incubator subdir; `applyDefer` stamps `deferred_until`; `applyEdit` returns the draft path for `$EDITOR`. Full procedure: `./apply-reflections-procedure.md` §[KFM-CANDIDATE].
-
-## [INSTINCT]
-
-Atomic instinct units emitted by `design-reflector` (and surfaced from `/gdd:extract-learnings`) appear as a distinct proposal class, alongside `[INCUBATOR]` and `[KFM-CANDIDATE]`. Each unit is a fenced `yaml` block under the reflector's `## Atomic instincts` section, shaped per `reference/instinct-format.md` (`id`, `trigger`, `confidence`, `domain`, `scope`, `project_id`, `source`, `cycles_seen`, `first_seen`, `last_seen`, plus a short body). A unit is a proposal, never a stored fact - nothing lands until the user accepts it.
-
-Mirror the `[INCUBATOR]` flow:
-
-1. Discover the units: parse every `yaml` block under `## Atomic instincts` in the reflections file. Skip malformed blocks (warn on stderr, keep going).
-2. For each unit: show the parsed `trigger`, `domain`, `confidence`, and body so the user sees what would be stored.
-3. Prompt: `(a) accept   (r) reject   (d) defer   (e) edit   (q) quit`.
-
-**Per-action behavior:**
-
-1. **accept** - call `scripts/lib/instinct-store.cjs` `add(unit, { scope, baseDir })` with the unit at its emitted `confidence`. The store owns de-duplication and `cycles_seen` bookkeeping. On success print `Stored instinct <id> (<domain>, confidence <n>).` and append `**Applied**: <date>` to the proposal block.
-2. **reject** - do not store the unit. Append `**Reviewed: rejected**` to the reflections file.
-3. **defer** - no-op; the unit re-surfaces next run. Append `**Reviewed: deferred**`.
-4. **edit** - let the user adjust `trigger`, `confidence`, or `domain`, then accept the edited unit through the same `add(...)` call. Default `scope: 'project'` (write `global` only when the unit's frontmatter says so); never edit `.design/instincts/instincts.json` directly, and promote to global via the separate gated `/gdd:instinct promote`.
-
-## Do Not
-
-- Do not apply any proposal without the user explicitly choosing `a` or `e`.
-- Do not modify source code files (`.ts`, `.tsx`, `.css`, `.js`) - only agent files, reference files, budget.json, discussant questions, global skills, and incubator drafts.
-- Do not re-run the reflector - this skill only applies existing proposals.
-- Do not bypass the scope guard or auto-flip Stage-1 - both are non-negotiable per D-05 / D-01.

package/dist/claude-code/.claude/skills/apply-reflections/apply-reflections-procedure.md

@@ -1,170 +0,0 @@
----
-name: apply-reflections-procedure
-type: heuristic
-version: 1.3.0
-phase: 30.5
-tags: [apply-reflections, proposal, frontmatter, reference, budget, question, global-skill, incubator, kfm-candidate]
-last_updated: 2026-05-21
----
-
-# Apply-Reflections - Per-Type Procedure
-
-Extracted from `skills/apply-reflections/SKILL.md` per Phase 28.5 D-10 (extract-then-link,
-never delete content). The orchestrator loop in `apply-reflections` (resolve file → parse →
-review loop → summary) stays in the SKILL. The per-proposal-type apply logic below moves
-here because it is content-class methodology, not workflow.
-
-## Apply Logic by Proposal Type
-
-After the user chooses `a` (apply) or `e` (edit-then-apply) in the review loop, branch by
-the proposal's bracketed type tag.
-
-### [FRONTMATTER]
-
-1. Extract agent name from Change field (e.g., `agents/design-verifier.md`)
-2. Read the agent file
-3. Find the frontmatter line matching the field being changed
-4. Use Edit tool to update the specific line
-5. Append `**Applied**: <date>` to the proposal in reflections file
-
-### [REFERENCE]
-
-1. Extract target file path from Change field (e.g., `reference/heuristics.md`)
-2. If file exists: append the drafted text using Edit tool
-3. If file doesn't exist: create it with a minimal header + the drafted text using Write tool
-4. Append `**Applied**: <date>` to proposal in reflections file
-
-### [BUDGET]
-
-1. Read `.design/budget.json`
-2. Locate the key path from the Change field (e.g., `design-verifier.per_run_cap_usd`)
-3. Update the value
-4. Write updated JSON back to `.design/budget.json`
-5. Append `**Applied**: <date>` to proposal in reflections file
-
-### [QUESTION]
-
-1. Read `agents/design-discussant.md`
-2. Find the question text specified in the Change field
-3. If pruning: remove the question lines using Edit tool
-4. If rewording: replace the question text using Edit tool
-5. Append `**Applied**: <date>` to proposal in reflections file
-
-### [GLOBAL-SKILL]
-
-1. Extract target filename from Change field (e.g., `design-color-conventions.md`)
-2. Ensure `~/.claude/gdd/global-skills/` directory exists (create with `mkdir -p` if not)
-3. If target file exists: append new content using Edit tool (add a `---` separator first)
-4. If target file doesn't exist: create with header + content using Write tool:
-
-   ```markdown
-   # <Topic> Conventions (Global)
-   *Promoted from project: <project-name>, cycle: <cycle-slug>*
-
-   <content>
-   ```
-
-5. Print: "Global skill written to ~/.claude/gdd/global-skills/<name>.md - auto-loads in all future gdd sessions"
-6. Append `**Applied**: <date>` to proposal in reflections file
-
-### [INCUBATOR]
-
-Incubator drafts come from `scripts/lib/incubator-author.cjs` (Phase 29-04). They live at
-`.design/reflections/incubator/<slug>/` and contain `manifest.json` + `DRAFT.md` + (optional) `ORIGIN.md`.
-
-Use `scripts/lib/apply-reflections/incubator-proposals.cjs` for all actions.
-
-**Discovery + render** (once per cycle):
-
-1. Call `discoverIncubatorDrafts()` → `Array<Draft>`. Skip malformed entries silently (already warned on stderr by the helper).
-2. For each draft: call `renderProposal(draft)` and print the returned markdown block. The user sees a header (slug + kind), a diff vs the nearest existing artifact (or "net-new"), an Origin section listing capability-gap signals, and the full draft body.
-3. Prompt: `(a) accept   (r) reject   (d) defer   (e) edit   (q) quit`.
-
-**Per-action behavior:**
-
-1. **accept** - call `applyAccept(draft, { registryPath, repoRoot })`.
-   - The helper calls `validateScope(draft.target_path)` from `scripts/validate-incubator-scope.cjs` **before** any write. Out-of-scope paths throw and the registry stays untouched. This is the non-bypassable scope guard (D-05).
-   - On success: target artifact written, `reference/registry.json` appended with `{ slug, path, added, origin: 'incubator' }`, incubator subdir removed last (T-29.05-04 - partial-failure leaves draft retryable).
-   - Print: "Accepted - promoted to <target_path>; registered."
-   - Append `**Applied**: <date>` to the proposal block.
-
-2. **reject** - call `applyReject(draft)`. Only the incubator subdir is removed; registry is untouched. Append `**Reviewed: rejected**` to the reflections file.
-
-3. **defer** - no-op. Print "Deferred - draft re-surfaces next run." Append `**Reviewed: deferred**`.
-
-4. **edit** - call `applyEdit(draft)` (uses `$EDITOR` or the `editorCmd` array option). On clean exit, the helper reloads the draft and the caller re-runs `renderProposal` + the prompt. On non-zero exit, the original draft is preserved unchanged.
-
-**Stage-1 gate (D-01 - no auto-flip):**
-
-1. At the start of the cycle, call `checkStage1Gate({ gateSpecPath, statePath, registryPath })`. The call is **read-only** - never mutates state. The returned `{ thresholdMet, summary, optInRecorded }` is informational.
-2. If `thresholdMet && !optInRecorded`, surface a one-time prompt:
-   ```
-   Stage-1 capability-gap authoring threshold met: <summary>
-   Enable incubator-draft promotion?  (y/N)
-   ```
-3. **Only on explicit `y`**, call `recordOptIn({ statePath, confirmedBy })`. The function is idempotent - a second call detects the existing record and returns `{ alreadyRecorded: true }`. Never call it on any other input.
-
-**Why this is gated.** The `[INCUBATOR]` proposal class can write executable surface (agents + skills) into the plugin runtime. Both Phase 29 D-01 (no auto-flip) and D-05 (scope guard) exist because that surface has integration-test and security implications that exceed reflector autonomy. `validateScope` keeps the file landing zone confined to `agents/<slug>.md` or `skills/<slug>/SKILL.md`. The Stage-1 gate keeps the *whether* of opting in to incubator authoring under explicit user control even after the data threshold says we have enough signal.
-
-**Bandit-fairness gate on `accept` (Phase 29 Plan 06 / CONTEXT D-04).**
-
-When the `accept` action promotes an incubator draft, the bandit-router arms for the freshly-promoted agent/skill MUST be bootstrapped with `prior_class: 'promoted_incubator'`. This invokes a conservative `Beta(2, 8)` bootstrap prior (posterior mean 0.2) instead of the optimistic Phase 23.5 informed prior - the bandit-fairness gate IS the staging mechanism (D-04: no separate two-step ratify split). The conservative prior suppresses preferential selection until ~8-10 successful pulls accumulate.
-
-Call shape (whether eagerly invoked on promotion or via first-pull lazy bootstrap):
-
-```javascript
-const bandit = require('./scripts/lib/bandit-router.cjs');
-// Per arm bootstrapped for the freshly-promoted agent:
-bandit.update({
-  agent: '<promoted-slug>',
-  bin: '<touches-bin>',
-  tier: '<chosen-tier>',
-  reward: <bernoulli>,
-  prior_class: 'promoted_incubator',  // Phase 29 Plan 06 / D-04 — Beta(2,8) staging
-});
-// Or for the delegate-aware case (Plan 27-07):
-bandit.updateWithDelegate({
-  agent: '<promoted-slug>',
-  bin: '<touches-bin>',
-  tier: '<chosen-tier>',
-  delegate: '<peer-cli-or-none>',
-  reward: <bernoulli>,
-  prior_class: 'promoted_incubator',
-});
-```
-
-Omitting `prior_class` reverts to Phase 23.5 informed-prior bootstrap (non-breaking). The reward math is unchanged - `prior_class` only affects bootstrap.
-
-### [KFM-CANDIDATE]
-
-Known-failure-mode catalogue proposals come from `scripts/lib/reflector-kfm-proposer.cjs` (Phase 30.5-03 D-05). They live at `.design/reflections/incubator/kfm-<slug>/CATALOGUE-ENTRY.md` and contain a single fenced ```yaml block pre-filled with the Phase 30.5 schema-v2 11-field shape (`id` + `pattern` + `diagnosis` + `remedy` + `severity` + `propose_report` + `symptom` + `root_cause` + `fix` + `related_phases` + `first_observed_cycle`). Two of those — `pattern` and `fix` — are `TODO:` placeholders the reflector cannot infer; the user fills them via the **edit** action before accepting.
-
-Two upstream signals share this draft surface (D-06):
-- `capability_gap` clusters of size ≥3 with no existing-entry match (Phase 29-03 aggregator + `failure-mode-matcher.match()`).
-- `kfm-candidate` events from the Phase 30.5-03 Task 2 authority-watcher whitelist (D-06 - single events bypass the ≥3 gate).
-
-Use `scripts/lib/reflector-kfm-proposer.cjs` for all actions:
-
-**Discovery + render** (once per cycle):
-
-1. Glob `.design/reflections/incubator/kfm-*/CATALOGUE-ENTRY.md` → list pending KFM drafts.
-2. For each draft: read the body, show the origin header (source, parent event ids OR article url) + the proposed yaml block.
-3. Prompt: `(a) accept   (r) reject   (d) defer   (e) edit   (q) quit`.
-
-**Per-action behavior:**
-
-1. **accept** - call `applyAccept(draftPath, { repoRoot })`.
-   - The helper re-stamps the proposed `id` with the next available `KFM-NNN` from the catalogue (avoids collisions when multiple drafts promote in the same run).
-   - Appends a `### KFM-NNN — <symptom heading>` section into `reference/known-failure-modes.md` with the yaml block intact.
-   - Appends a `reference/registry.json` entry: `{ name: 'known-failure-modes/kfm-NNN', path: 'reference/known-failure-modes.md', type: 'failure-mode', phase: 30.5, origin: 'incubator-kfm', added: '<ISO date>' }`.
-   - Removes the incubator subdir LAST (partial-failure leaves the draft retryable).
-   - Print: "Accepted - promoted to KFM-NNN in reference/known-failure-modes.md."
-   - Append `**Applied**: <date>` to the proposal entry (when surfaced from a reflections file).
-
-2. **reject** - call `applyReject(draftPath)`. Only the incubator subdir is removed; catalogue + registry untouched. Print: "Rejected - draft removed."
-
-3. **defer** - call `applyDefer(draftPath, { deferredUntil })` where `deferredUntil` is an ISO date (default: today + 30d). The helper stamps `deferred_until: <ISO>` into the draft body. Print: "Deferred - draft re-surfaces next run."
-
-4. **edit** - call `applyEdit(draftPath)` which returns the draft path. The caller opens `$EDITOR` on the path; on clean exit, re-discover the draft and re-prompt. Typical edits: replace `pattern: 'TODO: ...'` with a conservative regex, replace `fix: 'TODO: ...'` with a step-by-step user-runnable remedy, set `severity` if `medium` default is wrong.
-
-**Why this is gated.** `reference/known-failure-modes.md` feeds Phase 30's `triage-matcher.cjs` BEFORE the consent prompt - a bad entry could mute legitimate issue reports. The user-review gate is non-negotiable (D-05). The proposer is strictly proposal-only; the canonical catalogue only changes via the accept action.

package/dist/claude-code/.claude/skills/audit/SKILL.md

@@ -1,79 +0,0 @@
----
-name: gdd-audit
-description: "Run a design audit by spawning design-auditor, design-integration-checker, and (optionally) design-verifier + design-reflector agents, then printing a consolidated 7-pillar score summary. Use when the user wants to score the current design, retroactively verify a completed cycle, or quickly re-check after a fix. Activates for requests involving scoring an existing design, retroactively reviewing quality, or re-checking after a fix."
-argument-hint: "[--retroactive] [--quick] [--no-reflect]"
-tools: Read, Write, Task, Glob, Bash
----
-
-# /gdd:audit
-
-Wraps the existing `design-auditor`, `design-verifier`, and `design-integration-checker` agents - no new auditor logic here. Parses flags, spawns the right combination, prints summary.
-
-For the 7-pillar scoring rubric this skill aggregates, see `../../reference/audit-scoring.md`. For the shared design-quality pillar set that frames the score categories, see `../../reference/shared-preamble.md`.
-
-## Modes
-
-### Default
-Spawn `design-auditor` (7-pillar scoring 1–4) in parallel with `design-integration-checker`. After both finish, read `.design/DESIGN-AUDIT.md` and `.design/DESIGN-INTEGRATION.md` and print a consolidated summary (scores + top 3 findings each).
-
-After the auditor and integration checker complete, check if `.design/learnings/` exists and contains at least one `.md` file. If so - and unless `--no-reflect` is passed - spawn `design-reflector` for the current cycle. Append the reflection proposal count to the audit summary: "Reflection: N proposals → review with `/gdd:apply-reflections`".
-
-### `--retroactive`
-Spawn `design-verifier` with cycle-span scope. Verifier reads all tasks completed in the current cycle (from STATE.md `<completed_tasks>` list for the active `cycle:` ID) and uses `DESIGN-PLAN.md` goals as the reference baseline for what "should have been done." Output: `.design/DESIGN-VERIFICATION.md` with per-task pass/fail.
-
-### `--quick`
-Run only `design-auditor` (skip `design-integration-checker`). Faster health check when integration isn't the concern.
-
-## Steps
-
-1. Parse args for `--retroactive`, `--quick`, and `--no-reflect`.
-2. Verify `.design/STATE.md` exists; abort if not (suggest `/gdd:new-project`).
-3. Spawn the appropriate agents (Task tool). Default and `--retroactive` spawn two agents; `--quick` spawns one.
-4. Wait for completion, then read each output file and print a summary:
-   - Auditor scores per pillar
-   - Integration check pass/fail
-   - Verifier pass/fail per task (retroactive mode)
-5. **Reflection step** (default + retroactive modes only, skipped with `--no-reflect` or `--quick`):
-   - Check if `.design/learnings/` exists and has ≥1 `.md` file
-   - If yes: spawn `design-reflector` for the current cycle slug (read from STATE.md)
-   - After completion: count proposal types and append to summary
-6. Recommend next action based on findings (e.g., "Score 2/4 on typography - run `/gdd:discuss typography` to gather decisions").
-
-## Registered Audit Agents
-
-| Agent | Trigger | Output |
-|---|---|---|
-| `design-auditor` | Default, retroactive | `.design/DESIGN-AUDIT.md` |
-| `design-integration-checker` | Default, retroactive | `.design/DESIGN-INTEGRATION.md` |
-| `design-verifier` | `--retroactive` only | `.design/DESIGN-VERIFICATION.md` |
-| `design-reflector` | Default + retroactive when learnings exist | `.design/reflections/<slug>.md` |
-
-## Do Not
-
-- Do not modify source files.
-- Do not rerun stages; this is a read-only audit.
-
-## Step 7 - Update notice (post-closeout surface)
-
-After the consolidated audit summary has been printed (and any reflection-proposal count appended), emit the plugin-update banner if one is present:
-
-```bash
-[ -f .design/update-available.md ] && cat .design/update-available.md
-```
-
-Written by `hooks/update-check.sh`; suppressed mid-pipeline and when the latest release is dismissed.
-
-## Rationalizations - Thought to Reality
-
-The excuses an agent reaches for to skip or thin out an audit, and the drift each one misses:
-
-| Thought | Reality |
-|---------|---------|
-| "The audit passed last cycle, I can skip it this cycle." | Per-cycle audit catches drift the prior pass couldn't see; a skipped review is exactly where regressions accumulate unnoticed. |
-| "`--quick` is fine, integration isn't the concern here." | Dropping the integration-checker hides orphaned decisions - wiring breaks even when the 7-pillar score looks healthy. |
-| "I can eyeball the scores instead of spawning the auditor." | The auditor's rubric scores seven pillars consistently; an eyeballed review drifts toward whatever the agent already believes. |
-| "Reflection proposals are optional polish, skip the reflector." | The reflector turns this cycle's learnings into next-cycle improvements; skipping it lets the same mistakes repeat. |
-| "I'll modify the source while I'm in here fixing findings." | Audit is read-only by contract; editing source mid-audit invalidates the very scores you're producing. |
-| "Retroactive mode is overkill for a finished cycle." | Retroactive verification is the only check on tasks that shipped without per-task verify - skipping it leaves a completed cycle unaudited. |
-
-## AUDIT COMPLETE

package/dist/claude-code/.claude/skills/bandit-status/SKILL.md

@@ -1,94 +0,0 @@
----
-name: gdd-bandit-status
-description: "Surface read-only per-(agent, bin, delegate) bandit posterior snapshot - alpha/beta/mean/stddev/count/last-used per arm. Phase 27.5 (v1.27.5) diagnostic. Use when investigating 'why did the bandit pick tier X for agent Y?' or when verifying posterior convergence after enabling adaptive_mode: full."
-argument-hint: ""
-tools: Read, Bash
----
-
-# gdd-bandit-status
-
-## Role
-
-You are a deterministic, read-only diagnostic skill. You do not spawn agents and do not modify the posterior. You read `.design/telemetry/posterior.json` (path declared by `scripts/lib/bandit-router.cjs`'s `DEFAULT_POSTERIOR_PATH`), aggregate per-`(agent, bin, delegate, tier)` arm state, and emit a single Markdown table. Read-only per Phase 27.5 D-11 - to reset, use `/gdd:bandit-reset` (Phase 23.5). See `./reference/bandit-integration.md` for setup, interpretation, and convergence guidance.
-
-## Invocation Contract
-
-- **Input**: none.
-- **Output**: a Markdown bandit-status table to stdout. The table is the entire output.
-
-## Procedure
-
-### 1. Locate the posterior file
-
-Read `.design/telemetry/posterior.json`. Missing → emit empty-state message:
-
-```
-## Bandit Posterior Snapshot
-
-No posterior data yet — run a few pipeline cycles with `adaptive_mode: full` first.
-
-No posterior data found at `.design/telemetry/posterior.json`.
-
-Possible reasons:
-- `adaptive_mode` is `static` or `hedge` (bandit silent — see `.design/budget.json`).
-- No spawns have fired since Phase 27.5 wiring landed.
-- Posterior was cleared via `/gdd:bandit-reset`.
-
-See `reference/bandit-integration.md` for setup guidance.
-```
-
-Skip to Section 4 (Record). Parse failure (truncated/corrupted) → emit `Posterior file exists but is unparseable. Run /gdd:bandit-reset to start fresh, or restore from a backup.`
-
-### 2. Parse the posterior
-
-Schema:
-
-```json
-{
-  "schema_version": "1.0.0",
-  "generated_at": "<ISO>",
-  "arms": [{ "agent": "...", "bin": "...", "tier": "...", "delegate": "...", "alpha": N, "beta": N, "last_used": "...", "count": N }]
-}
-```
-
-The `delegate` field is optional - absent = Phase 23.5 legacy slice (rendered as `-` in the table).
-
-### 3. Render the table
-
-Compute per arm: `mean = alpha / (alpha + beta)` (3 decimals), `stddev = sqrt(alpha*beta / ((alpha+beta)^2 * (alpha+beta+1)))` (3 decimals).
-
-Sort by `(agent ASC, bin ASC, delegate ASC where '-' first, tier ASC opus<sonnet<haiku, last_used DESC)`. Group by agent for readability.
-
-Emit:
-
-```
-## Bandit Posterior Snapshot
-
-Per-(agent, bin, delegate, tier) posterior state. Read-only — to reset use `/gdd:bandit-reset` (Phase 23.5).
-
-Posterior file: `.design/telemetry/posterior.json` (last updated: <generated_at>)
-Total arms: <count>
-
-| Agent | Bin | Delegate | Tier | Alpha | Beta | Mean | Stddev | Count | Last Used |
-|-------|-----|----------|------|-------|------|------|--------|-------|-----------|
-| ...   | ... | ...      | ...  | ...   | ...  | ...  | ...    | ...   | ...       |
-
-> Mean = alpha / (alpha + beta). Stddev = sqrt(alpha*beta / ((alpha+beta)^2 * (alpha+beta+1))).
-> Delegate '-' = Phase 23.5 legacy slice (equivalent to 'none').
-> See `reference/bandit-integration.md` for interpretation.
-> Read-only — use `/gdd:bandit-reset` to clear posterior state.
-```
-
-Precision: alpha/beta 2 decimals; mean/stddev 3 decimals; count integer; `last_used` truncated to minute (`YYYY-MM-DDTHH:MM`); null `last_used` renders `-`.
-
-After the table, surface a per-`(agent, bin)` best-arm summary: for each unique pair, identify highest-mean arm (tie-broken by `count` DESC) - answers "why did the bandit pick tier X?" at a glance.
-
-### 4. Record
-
-Append one JSONL line to `.design/skill-records.jsonl`: `{"skill":"gdd-bandit-status","ts":"<ISO>","arms_seen":<count>,"posterior_present":<bool>}`. Skill writes ONLY to skill-records.jsonl (telemetry); never touches the posterior.
-
-## Cross-references
-
-- `./reference/bandit-integration.md` - operator guide; interpretation patterns.
-- `scripts/lib/bandit-router.cjs` (Phase 23.5) - posterior shape, `DEFAULT_POSTERIOR_PATH`, `loadPosterior()`.
-- `scripts/lib/bandit-router/integration.cjs` (27.5-01), `hooks/budget-enforcer.ts` (27.5-02), `scripts/lib/session-runner/index.ts` (27.5-03), `scripts/lib/bandit-arbitrage.cjs` (27.5-04), `/gdd:bandit-reset` (Phase 23.5) - only surface that mutates the posterior.

package/dist/claude-code/.claude/skills/benchmark/SKILL.md

@@ -1,65 +0,0 @@
----
-name: gdd-benchmark
-description: "Harvest and synthesize per-component design benchmarks from 18 design systems and produce canonical component specs at `reference/components/<name>.md`. Use when adding a new component spec, running a benchmark wave, listing corpus coverage, or refreshing a spec after a design-system version bump."
-argument-hint: "<component> | --wave <N> | --list | --refresh <component>"
-tools: Read, Write, Bash, Grep, Glob, Task, WebFetch
----
-
-# /gdd:benchmark
-
-Harvest per-component design knowledge from 18 design systems and synthesize canonical
-specs at `reference/components/<name>.md`. The 18-source corpus + fallback chain lives in
-`../../connections/design-corpora.md`. Per-skill output discipline + completion-marker
-conventions are at `../../reference/shared-preamble.md#output-contract-reminders`.
-
-## Invocation Modes
-
-| Invocation | Action |
-|------------|--------|
-| `/gdd:benchmark <component>` | Harvest + synthesize a single component |
-| `/gdd:benchmark --wave <N>` | Run a full wave (1 = Inputs, 2 = Containers, etc.) |
-| `/gdd:benchmark --list` | Show corpus coverage - which specs exist, which are pending |
-| `/gdd:benchmark --refresh <component>` | Re-harvest a spec (for design-system version bumps) |
-
-## Single-Component Flow (`/gdd:benchmark <component>`)
-
-1. **Check if spec exists** - `Glob("reference/components/<component>.md")`. If found and `--refresh` was not passed, confirm before overwriting.
-
-2. **Harvest** - spawn `component-benchmark-harvester` with required-reading `@connections/design-corpora.md`, target component, raw-output path `.planning/benchmarks/raw/<component>.md`, and a salvage hint to `.planning/research/impeccable-salvage/`. Acceptance: raw harvest exists with ≥4 source sections.
-
-3. **Synthesize** - spawn `component-benchmark-synthesizer` with required-reading `@.planning/benchmarks/raw/<component>.md`, `@reference/components/TEMPLATE.md`, `@reference/anti-patterns.md`. Output: `reference/components/<component>.md`. Also update `reference/components/README.md` (add entry in correct category). Acceptance: spec ≤350 lines, cites ≥4 systems, follows `TEMPLATE.md` sections, has a WAI-ARIA keyboard contract + a failing-example block.
-
-4. **Report** - print spec path, line count, systems cited.
-
-## Wave Mode (`/gdd:benchmark --wave <N>`)
-
-Read the wave definition from `reference/components/README.md` and run each component sequentially (not parallel - each harvest is network-bound and the raw files are large).
-
-| Wave | Components |
-|------|-----------|
-| 1 | button, input, select-combobox, checkbox, radio, switch, link, label |
-| 2 | card, modal-dialog, drawer, popover, tooltip, accordion, tabs |
-
-Print progress per component: `[N/total] harvesting <component>…`
-
-## List Mode (`/gdd:benchmark --list`)
-
-Read `reference/components/README.md` and diff against `reference/components/*.md` files. Print a table with columns: Component, Status, Wave, Lines.
-
-## Refresh Mode (`/gdd:benchmark --refresh <component>`)
-
-Same as single-component flow but skips the "already exists" guard. Use when a design system ships a breaking update to a component's spec.
-
-## Source List
-
-`../../connections/design-corpora.md` - 18 design systems with canonical URLs, licensing, and fallback chain (canonical → archive.org → Refero MCP → Pinterest MCP).
-
-## Output Artifacts
-
-| Artifact | Purpose |
-|----------|---------|
-| `.planning/benchmarks/raw/<component>.md` | Raw multi-source harvest (input only) |
-| `reference/components/<component>.md` | Canonical spec (distributed with plugin) |
-| `reference/components/README.md` | Corpus index (updated by synthesizer) |
-
-## BENCHMARK COMPLETE

package/dist/claude-code/.claude/skills/bootstrap-ds/SKILL.md

@@ -1,43 +0,0 @@
----
-name: gdd-bootstrap-ds
-description: "Bootstraps a design system for a GREENFIELD project that has none - no Figma, no tokens, no component library. Takes a brand input (primary color + optional secondary + tone tags + target framework) and emits a coherent OKLCH token system (color tints, modular type scale, 4pt/8pt spacing, radius + motion defaults) in 3 variants to pick from, then scaffolds proof components (button/input/card). Use at the start of a brand-new project, or when /gdd:discover finds no existing design system. Never invents a brand; never overwrites an existing DS."
-argument-hint: "[--primary <color>] [--secondary <color>] [--tone <tags>] [--framework web|native-ios|native-android|flutter]"
-user-invocable: true
-tools: Read, Write, Bash, Glob, Grep, AskUserQuestion, Task
----
-
-# /gdd:bootstrap-ds
-
-Greenfield design-system bootstrap. Closes the gap that GDD's `design-context-builder` assumes a design system already exists (in code or Figma) - a brand-new project has none. This skill is the **front door**: it collects a brand input and hands it to `agents/ds-generator.md`, which emits the token system + proof components per `reference/ds-bootstrap-rubric.md` (deterministic math in `scripts/lib/ds/token-scale.cjs`).
-
-## When to use
-
-- At the start of a brand-new project (no `tailwind.config`, no token file, no DS).
-- When `/gdd:discover` / `design-context-builder` reports **no existing design system** and the user opts to bootstrap one.
-
-If a design system **already exists**, do NOT run this - defer to `design-context-builder` (it maps the existing one). State that and stop.
-
-## Steps
-
-1. **Collect the brand input.** From flags, or via `AskUserQuestion` for anything missing:
-   - **primary** (required) - the brand color (hex / rgb / `oklch()`).
-   - **secondary** (optional) - a second brand color (emitted only if supplied - the rubric's ≤2-brand-colors rule).
-   - **tone tags** (optional) - `calm` / `corporate` / `editorial` / `playful` / `bold` (maps to the type ratio + chroma treatment).
-   - **target framework** (optional) - `web` (default) / `native-ios` / `native-android` / `flutter`. Detect from the project if absent (Phase 34 routing).
-2. **Delegate to `ds-generator`** (via `Task`): it resolves the primary to OKLCH, runs `token-scale.cjs` for **3 variants** (conservative / balanced / bold), and presents them.
-3. **Pick a variant.** Surface the 3 variants (primary `500`, type ratio, spacing baseline, radius, feel); the user picks ONE.
-4. **Emit + scaffold (proposal → confirm).** `ds-generator` emits the chosen token set (role-named CSS custom properties + the framework mapping) and scaffolds **button / input / card** as a coherence proof. Nothing is written to `src/` without confirmation.
-
-## Do Not
-
-- Do not invent a brand (no logomark, no typeface choice, no third brand color) - emit a token *system*, not an identity.
-- Do not overwrite an existing design system - defer to `design-context-builder`.
-- Do not add a color-conversion dependency - `token-scale.cjs` emits native CSS `oklch()`.
-
-## Output
-
-End with:
-
-```
-## BOOTSTRAP-DS COMPLETE
-```