devlyn-cli 3.2.1 → 3.3.1

package/AGENTS.md

@@ -38,11 +38,11 @@ Each skill's `SKILL.md` is the source of truth for flags and workflow. Do not du
 
 Engine roles: orchestrator = whichever CLI the user opened (this contract is symmetric with CLAUDE.md. If you are Codex orchestrating `/devlyn:resolve`, the phase machinery is mandatory regardless of task size; if you will not run it, say so explicitly and stop — do not silently degrade to ad-hoc execution); legacy executor (with optional separate worker/primary/pair profiles) uses the canonical skill's orchestrator-supported default, overridable per run with `--engine <name>` or durably via machine-local `.devlyn/engines.json` `{"executor": "<name>", "pair_judge_priority": ["<name>", ...]}`; PLAN is orchestrator-fixed and never inherits `--engine` or the executor pin; pair judge = first available OTHER engine, default for VERIFY whenever available; risk probes remain conditional and --no-pair opts out. Pins fail closed (`BLOCKED:<engine>-unavailable` / `BLOCKED:invalid-engine-config`); new engines plug in by shipping `_shared/adapters/<name>.md`. `/devlyn:engines` (no args) shows the role table + detected engines; `executor <name>` / `pair <name>,...` / `clear` manage the pins. The executor pin binds the orchestrator in plain conversation too, not only inside a skill run: when you would implement directly and executor is pinned to a non-default engine, route that work through the pin — via `/devlyn:resolve` (reads it at PHASE 0) or by delegating to that engine; no pin / no file → use the canonical skill's orchestrator-supported default. The pair-judge pin stays pipeline-scoped. `/devlyn:engines role` configures explicit engine/model/effort for supported worker and judge routes; `/devlyn:resolve --role-config <path>` overrides roles for one run. Canonical skills define precedence, evidence and unsupported-route errors; absent profiles preserve existing defaults.
 
-Conversational entry: Before loading a workflow or writing files, inspect the requested change and relevant callers/tests. Quoted phase names, past logs and skill-file paths are context, not an instruction to invoke that workflow; follow the current user request. Default to direct execution when inspection makes the requested behavior, affected boundaries and verification tractable in the current context, including bounded multi-file work. Preserve requested behavior, scope, explicit constraints and the executor pin; make scoped edits, run required checks, review the final diff, and report changes with evidence. One concise route/check explanation suffices; ordinary direct work needs no spec, pipeline state, phase workers or extra approval. For direct/full/queue work, the outer owner reads `references/task-completion.md` relative to the installed `devlyn:resolve` skill's directory: prospective task-branch ownership (linked worktree optional), scoped commit acceptance, then PR/merge and recoverable owned-resource cleanup after required archive/queue commits. Local-only/no-push wins; delivery pending/failure stays separate from product verification.
+Conversational entry: Before loading a workflow or writing files, inspect the requested change and relevant callers/tests. Quoted phase names, past logs and skill-file paths are context, not an instruction to invoke that workflow; follow the current user request. Default to direct execution when inspection makes the requested behavior, affected boundaries and verification tractable in the current context, including bounded multi-file work. Preserve requested behavior, scope, explicit constraints and the executor pin; make scoped edits, run required checks, review the final diff, and report changes with evidence. One concise route/check explanation suffices; ordinary direct work needs no spec, pipeline state, phase workers or extra approval. For direct/full/queue work, the outer owner reads `references/task-completion.md` relative to the installed `devlyn:resolve` skill's directory: prospective task-branch ownership (own linked worktree), scoped commit acceptance, then PR/merge and recoverable owned-resource cleanup after required archive/queue commits. Local-only/no-push wins; delivery pending/failure stays separate from product verification.
 
 Use full `/devlyn:resolve` automatically only when inspection identifies interacting requirements or verification work too complex to manage reliably in the current context, such as coupled durable state, concurrent ownership and failure recovery across boundaries. Name the concrete interaction or verification gap. Domain labels (security/auth/payment/persistence/concurrency/API), file count and a spec document alone do not trigger it. Investigate or clarify missing intent/access first; a pipeline cannot supply them. Direct work still requires risk-proportionate checks and independent review for consequential changes. Explicit resolve (even small), explicit spec-mode workflows and queue drains retain all canonical phases, independent verification, pins and failure handling. If the full-route condition emerges mid-edit, carry the existing delta into its scope/evidence; do not reset the baseline or reconfirm existing authorization.
 
-Full-route handoff + loop engineering: the orchestrating model — not the user — invokes the skills. `/devlyn:queue` fronts the intent queue (`docs/specs/queue.md`): no args = status, `add <intent>`, `drain` = serial unattended drain per the contract below. Work selected for full resolve is written to `docs/specs/<id>/spec.md` (always a spec file — the user's reviewed contract; free-form large assumes-and-logs instead of halting, zero-scope-signal goals still halt to ideate), summarized once for user review, then run hands-free via `--spec`. Outer loop per task: before the first full run, commit the scoped queue-item delta (when applicable) and linked spec bundle; findings-backed verdicts (NEEDS_WORK, verify/build-gate exhaustion) get a committed spec amendment before re-run, max 3 iterations; infrastructure / invalid-input / engine-availability / implement-empty BLOCKED verdicts surface immediately. Unattended queue drain (`docs/specs/queue.md`, strictly serial): assumptions may only take scope-narrowing, reversible, non-user-visible defaults — material ambiguity marks the item `[F] needs-review`; commit every terminal `[x]` / `[F]` transition before advancing. The canonical commit order is `config/skills/devlyn:resolve/references/outer-loop.md`.
+Full-route handoff + loop engineering: the orchestrating model — not the user — invokes the skills. `/devlyn:queue` fronts the intent queue (`docs/specs/queue.md`): no args = status, `add <intent>`, `drain` = serial unattended drain per the contract below. Work selected for full resolve is written to `docs/specs/<id>/spec.md` (always a spec file — the user's reviewed contract; free-form large assumes-and-logs instead of halting, zero-scope-signal goals still halt to ideate), summarized once for user review, then run hands-free via `--spec`. Outer loop per task: before the first full run, commit the scoped queue-item delta (when applicable) and linked spec bundle; findings-backed verdicts (NEEDS_WORK, BLOCKED:repair-budget-exhausted) get a committed spec amendment before re-run, max 3 iterations; infrastructure / invalid-input / engine-availability / implement-empty BLOCKED verdicts surface immediately. Unattended queue drain (`docs/specs/queue.md`, strictly serial): assumptions may only take scope-narrowing, reversible, non-user-visible defaults — material ambiguity marks the item `[F] needs-review`; commit every terminal `[x]` / `[F]` transition before advancing. The canonical commit order is `config/skills/devlyn:resolve/references/outer-loop.md`.
 
 ## Subtractive-First Editing
 

package/CLAUDE.md

@@ -51,7 +51,7 @@ Each skill's `SKILL.md` is the source of truth for its flags and workflow — do
 
 ### Conversational entry + full-route handoff
 
-Before loading a workflow or writing files, inspect the requested change and relevant callers/tests. Quoted phase names, past logs and skill-file paths are context, not an instruction to invoke that workflow; follow the current user request. Default to direct execution when inspection makes the requested behavior, affected boundaries and verification tractable in the current context, including bounded multi-file work. Preserve requested behavior, scope, explicit constraints and the executor pin; make scoped edits, run required checks, review the final diff, and report changes with evidence. One concise route/check explanation suffices; ordinary direct work needs no spec, pipeline state, phase workers or extra approval. For direct/full/queue work, the outer owner reads `references/task-completion.md` relative to the installed `devlyn:resolve` skill's directory: prospective task-branch ownership (linked worktree optional), scoped commit acceptance, then PR/merge and recoverable owned-resource cleanup after required archive/queue commits. Local-only/no-push wins; delivery pending/failure stays separate from product verification.
+Before loading a workflow or writing files, inspect the requested change and relevant callers/tests. Quoted phase names, past logs and skill-file paths are context, not an instruction to invoke that workflow; follow the current user request. Default to direct execution when inspection makes the requested behavior, affected boundaries and verification tractable in the current context, including bounded multi-file work. Preserve requested behavior, scope, explicit constraints and the executor pin; make scoped edits, run required checks, review the final diff, and report changes with evidence. One concise route/check explanation suffices; ordinary direct work needs no spec, pipeline state, phase workers or extra approval. For direct/full/queue work, the outer owner reads `references/task-completion.md` relative to the installed `devlyn:resolve` skill's directory: prospective task-branch ownership (own linked worktree), scoped commit acceptance, then PR/merge and recoverable owned-resource cleanup after required archive/queue commits. Local-only/no-push wins; delivery pending/failure stays separate from product verification.
 
 Use full `/devlyn:resolve` automatically only when inspection identifies interacting requirements or verification work too complex to manage reliably in the current context, such as coupled durable state, concurrent ownership and failure recovery across boundaries. Name the concrete interaction or verification gap. Domain labels (security/auth/payment/persistence/concurrency/API), file count and a spec document alone do not trigger it. Investigate or clarify missing intent/access first; a pipeline cannot supply them. Direct work still requires risk-proportionate checks and independent review for consequential changes. Explicit resolve (even small), explicit spec-mode workflows and queue drains retain all canonical phases, independent verification, pins and failure handling. If the full-route condition emerges mid-edit, carry the existing delta into its scope/evidence; do not reset the baseline or reconfirm existing authorization.
 
@@ -60,7 +60,7 @@ The user does not invoke skills manually; the orchestrating model does. For work
 1. Write the agreed contract to `docs/specs/<id>/spec.md` (+ `spec.expected.json` when mechanical verifications exist). Use a spec file for the selected work — a written spec is the user's reviewed contract; free-form large now assumes-and-logs instead of halting (zero-scope-signal goals still halt to ideate).
 2. Present a one-screen plan-contract summary — the user's single review checkpoint, BEFORE the pipeline starts.
 3. On go-ahead, commit the scoped queue-item delta (when applicable) and linked spec bundle, then invoke `/devlyn:resolve --spec <path>` hands-free to completion.
-4. **Per-task outer loop**: read the terminal verdict. Successful normal-run archive → outer-owner completion per the canonical task-completion reference, after any terminal queue commit. Verdicts backed by spec/verification findings (NEEDS_WORK, verify/build-gate exhaustion) → adjudicate the findings, amend the spec (recorded in the spec file; spec stays read-only inside a run), commit that scoped amendment, then re-invoke — at most 3 outer iterations, then surface with the findings trail. Infrastructure, invalid-input, engine-availability, and implement-empty BLOCKED verdicts are not spec-amendable: surface immediately. Every iteration re-enters through committed durable artifacts (spec, findings, run archive), never conversation memory.
+4. **Per-task outer loop**: read the terminal verdict. Successful normal-run archive → outer-owner completion per the canonical task-completion reference, after any terminal queue commit. Verdicts backed by spec/verification findings (NEEDS_WORK, BLOCKED:repair-budget-exhausted) → adjudicate the findings, amend the spec (recorded in the spec file; spec stays read-only inside a run), commit that scoped amendment, then re-invoke — at most 3 outer iterations, then surface with the findings trail. Infrastructure, invalid-input, engine-availability, and implement-empty BLOCKED verdicts are not spec-amendable: surface immediately. Every iteration re-enters through committed durable artifacts (spec, findings, run archive), never conversation memory.
 
 **Intent queue (unattended drain)**: the user stacks intents in `docs/specs/queue.md` (ordered checklist; `/devlyn:queue` is the front-end — no args shows status, `add <intent>` appends, `drain` starts the serial drain). Drain strictly serially — per item: spec it, commit the scoped owner baseline, run the outer loop, mark `[x]` done or `[F]` blocked with reason, commit that terminal transition, then complete eligible delivery once. Retain PR/pending workspaces; unavailable separate safe placement stops the drain with remaining items pending. A product-blocked item otherwise never halts the queue. The queue entry is the user's go-ahead, so assume-and-log replaces the interactive checkpoint — but unattended assumptions may only take scope-narrowing, reversible, non-user-visible defaults. Material ambiguity (user-visible behavior, data/state semantics, new files/scripts/flags, implementation surface) → mark `[F] needs-review`, commit it, and move on. End the drain with a per-item product verdict, separate delivery status/URL/resume and assumptions report. The canonical commit order is `config/skills/devlyn:resolve/references/outer-loop.md`.
 

package/README.md

@@ -84,27 +84,14 @@ PLAN  →  IMPLEMENT  →  BUILD_GATE  →  CLEANUP  →  VERIFY (fresh subagent
 Common flags: `--engine claude|codex|omp` (default: the orchestrator-supported default), `--role-config <path>` (one-run worker/judge profiles, see below), `--bypass build-gate,cleanup`, `--pair-verify` (force pair-mode JUDGE in VERIFY), `--no-pair` (intentional solo VERIFY), `--risk-probes` / `--no-risk-probes`, `--perf` (per-phase timing).
 `--pair-verify` and `--no-pair` are mutually exclusive; using both stops with `BLOCKED:invalid-flags`.
 
-Free-form goals that ask for benchmark evidence, pair-evidence, risk-probe
-measurement, `solo<pair` proof, or solo-headroom work must include an
-actionable `solo-headroom hypothesis` naming the visible behavior `solo_claude`
-is expected to miss plus a backticked observable command; the backticked line
-itself must contain `miss` and be framed as the command/observable that exposes it. Without that,
-`/devlyn:resolve` stops with `BLOCKED:solo-headroom-hypothesis-required` and
-points you to `/devlyn:ideate` instead of inventing a weak hypothesis.
-Free-form goals that add or run a new unmeasured benchmark, shadow fixture,
-golden fixture, risk-probe, or pair-evidence candidate must also include
-`solo ceiling avoidance`, mention `solo_claude`, and name the concrete
-difference from rejected or solo-saturated controls such as `S2`-`S6`; without
-that, `/devlyn:resolve` stops with `BLOCKED:solo-ceiling-avoidance-required`.
-
-Accepted tasks default to owner completion: scoped commit → push → PR → normal
-protected merge → recoverable cleanup of prospectively owned resources. Set
-`git config --local devlyn.completionMode pr` to stop at a PR, or `auto` to restore
-the default; `task-complete.py complete --mode auto|pr` overrides one task.
-Local-only/no-push instructions take precedence. Linked worktrees are optional;
-existing branches cannot be adopted. Full runs require successful archive before
-delivery; direct tasks use their actual checks and root acceptance. Verify-only
-never publishes. Pending checks or unsupported merge policy retain resources and
+Each task gets its own linked worktree; accepted tasks default to scoped commit → push → PR
+→ merge when the repository allows it (otherwise the PR waits for a person). After the
+PR merges, the worktree and branch its session released are cleaned; anything in use is
+kept. Set `git config --local devlyn.completionMode pr` to stop at the PR; `task-complete.py complete --mode auto|pr`
+overrides one task. Local-only/no-push instructions take precedence; existing
+branches cannot be adopted. Full runs require successful archive before delivery;
+direct tasks use their actual checks and root acceptance. Verify-only never
+publishes. Pending checks or unsupported merge policy retain resources and
 report a receipt-based resume command separately from product verification.
 See [task completion](config/skills/devlyn:resolve/references/task-completion.md)
 for allocation, acceptance, writer cessation and recovery.
@@ -157,74 +144,6 @@ Pin roles durably with `/devlyn:engines` (no args shows the role table + detecte
 
 If an engine is absent when explicitly selected by flag, pin, or role profile, or OTHER engine is absent under `--pair-verify`, the harness stops with `BLOCKED:<engine>-unavailable` and prints setup guidance. Automatic VERIFY absence is a reported solo route. Use `--no-pair` only when intentionally accepting solo VERIFY; use `--no-risk-probes` only when intentionally disabling automatic high-risk probes.
 
-### Benchmark score runs
-
-Use the benchmark CLI when a change claims `solo_claude < pair`. The score-focused runners print the run id, startup gate lines, blind-judge score tables, fixture pair margins, average pair margin, wall-time ratio, and failure reasons:
-
-```bash
-npx devlyn-cli benchmark headroom --min-fixtures 3 F16-cli-quote-tax-rules F23-cli-fulfillment-wave F25-cli-cart-promotion-rules
-npx devlyn-cli benchmark recent
-npx devlyn-cli benchmark recent --out-md /tmp/devlyn-recent-benchmark.md
-npx devlyn-cli benchmark frontier --out-md /tmp/devlyn-pair-frontier.md
-npx devlyn-cli benchmark audit --out-dir /tmp/devlyn-benchmark-audit
-npx devlyn-cli benchmark audit-headroom --out-json /tmp/devlyn-headroom-audit.json
-npx devlyn-cli benchmark pair --min-fixtures 3 --max-pair-solo-wall-ratio 3 F16-cli-quote-tax-rules F23-cli-fulfillment-wave F25-cli-cart-promotion-rules
-```
-
-`benchmark recent` prints a compact, wrap-safe snapshot of the current local
-pair evidence: status counts, pair-lift aggregates, and one card per passing
-pair-evidence fixture. It intentionally avoids wide Markdown tables, so the
-same output stays readable in narrow terminals, PR comments, and release notes.
-`benchmark frontier` also prints a stdout score summary for existing complete pair
-evidence rows, including pair arm, trigger reasons, average/minimum pair margin,
-and wall ratio, plus row-level verdicts even when `--out-json` or `--out-md`
-writes an artifact. Markdown frontier artifacts include a `Triggers` column.
-Full-pipeline pair gate artifacts record `require_hypothesis_trigger` in JSON
-and include a Markdown `Hypothesis trigger` column, so strict regenerated
-evidence shows whether each row carried `spec.solo_headroom_hypothesis`.
-`benchmark audit` is the provider-free release/handoff guard: it writes
-`audit.json` with the frontier summary, artifact map, and compact trigger-backed verdict-bearing `pair_evidence_rows`
-(each row carries `pair_trigger_eligible: true`, non-empty `pair_trigger_reasons`, `pair_trigger_has_canonical_reason: true`, and `pair_trigger_has_hypothesis_reason`; the audit fails rows missing trigger reasons or missing actionable solo-headroom hypotheses in fixture `spec.md` whose observable command matches `expected.json`), runs the frontier with
-`--fail-on-unmeasured`, requires at least four fixtures with passing pair evidence,
-revalidates frontier `verdict: PASS`, zero unmeasured candidates, and revalidates `pair_mode: true`,
-the default 5-point pair margin, and 3x pair/solo wall ratio, then
-audits failed headroom results. The audit stdout also prints
-`headroom_rejections=...`, `pair_evidence_quality=...`,
-`pair_trigger_reasons=...`, `pair_evidence_hypotheses=...`, and
-`pair_evidence_hypothesis_triggers=...` handoff rows, plus
-`pair_trigger_historical_aliases=...` when archived evidence includes legacy
-trigger aliases and `pair_evidence_hypothesis_trigger_gaps=...` when documented
-hypotheses have not yet propagated into trigger reasons, with the rejected-fixture
-coverage counts plus actual minimum pair margin, maximum pair/solo wall ratio,
-and canonical trigger reason coverage plus row-match status.
-The compact evidence row count must match the frontier evidence count,
-`checks.frontier_stdout` records summary, aggregate, final-verdict, expected, printed score-row, trigger-visible row, and hypothesis-trigger-visible row counts,
-`checks.headroom_rejections` records child verdict plus unrecorded/unsupported counts,
-`checks.pair_evidence_quality` records the same quality thresholds from the compact rows,
-`checks.pair_trigger_reasons` records canonical/historical-alias/exposed/total trigger-reason row counts, fixture-level historical alias details, summary count, and row-match status,
-`checks.pair_evidence_hypotheses` records documented/total pair-evidence hypothesis row counts,
-and `checks.pair_evidence_hypothesis_triggers` records whether documented hypotheses also appear as `spec.solo_headroom_hypothesis` trigger reasons plus fixture-level gap details
-so incomplete or low-quality local score artifacts cannot inflate the claim.
-Add `--require-hypothesis-trigger` to turn those hypothesis-trigger gaps from
-archived-evidence WARN rows into release-blocking FAIL rows for newly
-regenerated pair evidence.
-
-```bash
-npx devlyn-cli benchmark audit --require-hypothesis-trigger --out-dir /tmp/devlyn-benchmark-audit-strict
-```
-
-Historical trigger aliases are only reported for archived artifact review; new
-current pair-evidence gates fail historical-only or unknown trigger reasons and
-require at least one canonical `pair_trigger.reasons` entry.
-`benchmark audit-headroom` fails if an active failed headroom fixture is missing
-from both rejected registry and passing pair evidence.
-Headroom runs use the current claim gate: `bare <= 60`, `solo_claude <= 80`,
-and the default 5-point `bare`/`solo_claude` headroom margins before spending a pair arm.
-Add `--dry-run` to either score runner to validate args, fixture ids, minimum
-fixture count, and the replay command without running arms or judges. Dry-runs
-and lint prove wiring only; real score claims must cite the run id and fixture
-ids.
-
 ### Migration from earlier versions
 
 Reinstall with the new version to refresh instructions and skills for the selected
@@ -260,19 +179,6 @@ Earlier versions of devlyn-cli shipped 16+ slash commands. The iter-0034 Phase 4
 
 ---
 
-## Auto-Activated Skills
-
-These activate automatically — no commands needed. They shape how Claude thinks during relevant tasks.
-
-| Skill | Activates During |
-|---|---|
-| `root-cause-analysis` | Debugging — enforces 5 Whys, evidence standards |
-| `code-review-standards` | Reviews — severity framework, approval criteria |
-| `ui-implementation-standards` | UI work — design fidelity, accessibility, responsiveness |
-| `code-health-standards` | Maintenance — dead code prevention, complexity thresholds |
-
----
-
 ## Optional Add-ons
 
 Selected during install. Run `npx devlyn-cli` again to add more.
@@ -292,6 +198,10 @@ Selected during install. Run `npx devlyn-cli` again to add more.
 | `devlyn:pencil-pull` | Pull Pencil designs into code |
 | `devlyn:pencil-push` | Push codebase UI to Pencil canvas |
 | `devlyn:reap` | Safely reap orphaned MCP / codex / Superset child processes |
+| `code-health-standards` | Maintainability standards — dead code, dependencies, complexity, naming |
+| `code-review-standards` | Severity framework and approval criteria for reviews |
+| `root-cause-analysis` | Evidence-first why-chain debugging |
+| `ui-implementation-standards` | UI quality bar — design tokens, accessibility, state coverage, responsive layout |
 
 </details>
 

package/bin/devlyn.js

@@ -13,8 +13,7 @@ const { updateInstructions, InstructionError } = require('./instructions');
 
 // The devlyn skill bundle installed into every skill-capable agent's loader
 // directory. Single source of truth so codex/omp/pi stay in lockstep — adding a
-// skill here installs it everywhere. Standards skills (code-*, root-cause-*,
-// ui-*) stay Claude-core-only, matching the pre-existing Codex bundle.
+// skill here installs it everywhere.
 const DEVLYN_CORE_SKILLS = ['devlyn:resolve', 'devlyn:ideate', 'devlyn:design-ui', 'devlyn:engines', 'devlyn:queue', '_shared'];
 const DEVLYN_SKILL_DIR_STAMP = '__DEVLYN_SKILL_DIR__';
 const DEVLYN_INSTALL_MARKER = '.devlyn-install.json';
@@ -169,6 +168,11 @@ const DEPRECATED_DIRS = [
   // upgrade so users only have them if they opt in via the interactive
   // installer (matches the pencil-pull / pencil-push pattern).
   'skills/devlyn:reap',
+  // 0221 Session 1: standards skills moved to optional-skills/ (opt-in).
+  'skills/code-health-standards',
+  'skills/code-review-standards',
+  'skills/root-cause-analysis',
+  'skills/ui-implementation-standards',
   // Deleted entirely on 2026-05-14 (devlyn:team-design-ui merged into
   // devlyn:design-ui; devlyn:design-system removed outright). Entries kept
   // so users who previously opted in get their stale copies purged on upgrade.
@@ -237,6 +241,10 @@ const OPTIONAL_ADDONS = [
   { name: 'devlyn:pencil-pull', desc: 'Pull Pencil designs into code with exact visual fidelity', type: 'local' },
   { name: 'devlyn:pencil-push', desc: 'Push codebase UI to Pencil canvas for design sync', type: 'local' },
   { name: 'devlyn:reap', desc: 'Safely reap orphaned MCP / codex / Superset child processes left behind by long Claude sessions', type: 'local' },
+  { name: 'code-health-standards', desc: 'Maintainability standards — dead code, dependencies, complexity, naming, hygiene', type: 'local' },
+  { name: 'code-review-standards', desc: 'Severity framework and approval criteria for reviewing a change', type: 'local' },
+  { name: 'root-cause-analysis', desc: 'Evidence-first why-chain debugging to the actionable root cause', type: 'local' },
+  { name: 'ui-implementation-standards', desc: 'UI quality bar — design tokens, WCAG 2.1 AA, state coverage, responsive layout', type: 'local' },
   // External skill packs (installed via npx skills add)
   { name: 'vercel-labs/agent-skills', desc: 'React, Next.js, React Native best practices', type: 'external' },
   { name: 'supabase/agent-skills', desc: 'Supabase integration patterns', type: 'external' },
@@ -660,7 +668,7 @@ function detectOtherCLIs() {
   return detected;
 }
 
-// Install devlyn:resolve + devlyn:ideate + devlyn:design-ui + _shared skills into a CLI's
+// Install DEVLYN_CORE_SKILLS into a CLI's
 // global skills directory (e.g. ~/.codex/skills/). Returns count of skills
 // copied. Skipped silently for CLIs without a skillsDir (e.g. cursor, copilot
 // at the time of writing — they don't have an analogous skill-loader).
@@ -722,7 +730,7 @@ function installAgentsForCLI(cliKey) {
   const instr = installInstructionsForCLI(cliKey);
   const skillsCopied = installSkillsForCLI(cliKey);
   if (skillsCopied > 0) {
-    log(`  → ${skillsCopied} skill${skillsCopied > 1 ? 's' : ''} installed (devlyn:resolve / devlyn:ideate / devlyn:design-ui / _shared)`, 'dim');
+    log(`  → ${skillsCopied} skill${skillsCopied > 1 ? 's' : ''} installed (${DEVLYN_CORE_SKILLS.join(' / ')})`, 'dim');
   }
   return instr || skillsCopied > 0;
 }
@@ -928,19 +936,11 @@ function installClaudeCore() {
   // Configure global Claude Code settings (~/.claude/settings.json)
   // Project settings may refer to this same file; merge the latest bytes.
   const globalSettings = readGlobalSettings();
-  let globalSettingsChanged = false;
-  if (!globalSettings.env.CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING) {
-    globalSettings.env.CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING = '1';
-    globalSettingsChanged = true;
-  }
   if (!globalSettings.env.ENABLE_PROMPT_CACHING_1H) {
     globalSettings.env.ENABLE_PROMPT_CACHING_1H = 'true';
-    globalSettingsChanged = true;
-  }
-  if (globalSettingsChanged) {
     if (!fs.existsSync(globalClaudeDir)) fs.mkdirSync(globalClaudeDir, { recursive: true });
     fs.writeFileSync(globalSettingsPath, JSON.stringify(globalSettings, null, 2) + '\n');
-    log('  → ~/.claude/settings.json (disabled adaptive thinking, enabled 1h prompt caching)', 'dim');
+    log('  → ~/.claude/settings.json (enabled 1h prompt caching)', 'dim');
   }
 
   log('\n✅ Claude Code config installed!', 'green');
@@ -1022,15 +1022,6 @@ function showHelp() {
   log('  npx devlyn-cli -y           Install Claude core without prompts');
   log('  npx devlyn-cli agents       Install agents for detected CLIs');
   log('  npx devlyn-cli agents all   Install agents for all supported CLIs');
-  log('  npx devlyn-cli benchmark    Run the resolve benchmark suite');
-  log('  npx devlyn-cli benchmark recent              Show compact recent benchmark results');
-  log('  npx devlyn-cli benchmark frontier            Show pair candidate frontier scores/triggers without providers');
-  log('  npx devlyn-cli benchmark audit               Audit pair evidence readiness');
-  log('  npx devlyn-cli benchmark audit-headroom      Audit failed headroom results');
-  log('  npx devlyn-cli benchmark headroom <fixtures...>  Score bare vs solo_claude headroom');
-  log('  npx devlyn-cli benchmark pair <fixtures...>      Score solo_claude vs pair path');
-  log('  npx devlyn-cli benchmark --bless         If ship-gate passes, promote baseline');
-  log('  npx devlyn-cli benchmark --dry-run       Validate suite setup without model invocation');
   log('  npx devlyn-cli --help       Show this help\n');
   log('Optional skills (select during install):', 'green');
   OPTIONAL_ADDONS.filter((a) => a.type === 'local').forEach((skill) => {
@@ -1051,170 +1042,6 @@ function showHelp() {
   log('');
 }
 
-function showBenchmarkHelp() {
-  log('Usage:', 'green');
-  log('  npx devlyn-cli benchmark [suite] [options] [fixtures...]');
-  log('  npx devlyn-cli benchmark recent [options]');
-  log('  npx devlyn-cli benchmark frontier [options]');
-  log('  npx devlyn-cli benchmark audit [options]');
-  log('  npx devlyn-cli benchmark audit-headroom [options]');
-  log('  npx devlyn-cli benchmark headroom [options] <fixtures...>');
-  log('  npx devlyn-cli benchmark pair [options] <fixtures...>');
-  log('');
-  log('Score-focused runs:', 'green');
-  log('  recent   Show compact, wrap-safe recent benchmark results');
-  log('  frontier Show active rejected/evidence/unmeasured pair candidates, scores, and triggers without providers');
-  log('  audit     Fail on unmeasured pair candidates and invalid headroom rejections');
-  log('            Prints frontier score rows plus headroom and pair quality handoff rows');
-  log('  audit-headroom  Fail on active failed or unsupported headroom rejections');
-  log('  headroom  Score bare vs solo_claude before spending the pair arm');
-  log('  pair      Score solo_claude vs the selected pair path and print gate tables');
-  log('');
-  log('Shadow suite:', 'green');
-  log('  npx devlyn-cli benchmark suite --suite shadow --dry-run');
-  log('            Lists shadow tasks only; use headroom/pair with explicit S* ids for real measurement');
-  log('');
-  log('Examples:', 'green');
-  log('  npx devlyn-cli benchmark --dry-run F1-cli-trivial-flag');
-  log('  npx devlyn-cli benchmark recent');
-  log('  npx devlyn-cli benchmark recent --out-md /tmp/devlyn-recent-benchmark.md');
-  log('  npx devlyn-cli benchmark frontier --out-md /tmp/devlyn-pair-frontier.md');
-  log('  npx devlyn-cli benchmark audit --out-dir /tmp/devlyn-benchmark-audit');
-  log('  npx devlyn-cli benchmark audit-headroom --out-json /tmp/devlyn-headroom-audit.json');
-  log('  npx devlyn-cli benchmark headroom --min-fixtures 3 F16-cli-quote-tax-rules F23-cli-fulfillment-wave F25-cli-cart-promotion-rules');
-  log('  npx devlyn-cli benchmark pair --min-fixtures 3 --max-pair-solo-wall-ratio 3 F16-cli-quote-tax-rules F23-cli-fulfillment-wave F25-cli-cart-promotion-rules');
-  log('');
-}
-
-function showBenchmarkModeHelp(mode) {
-  if (mode === 'recent') {
-    log('Usage:', 'green');
-    log('  npx devlyn-cli benchmark recent [options]');
-    log('');
-    log('Options:', 'green');
-    log('  --out-json PATH');
-    log('  --out-md PATH');
-    log('  --fixtures-root PATH');
-    log('  --registry PATH');
-    log('  --results-root PATH');
-    log('  --max-width N  default: 92');
-    log('  --min-pair-margin N  default: 5');
-    log('  --max-pair-solo-wall-ratio N  default: 3');
-    log('');
-    log('Output:', 'green');
-    log('  Prints compact, wrap-safe benchmark status and pair-evidence cards without wide tables');
-    log('');
-    log('Example:', 'green');
-    log('  npx devlyn-cli benchmark recent');
-    log('  npx devlyn-cli benchmark recent --out-md /tmp/devlyn-recent-benchmark.md');
-    log('');
-    return;
-  }
-  if (mode === 'frontier') {
-    log('Usage:', 'green');
-    log('  npx devlyn-cli benchmark frontier [options]');
-    log('');
-    log('Options:', 'green');
-    log('  --out-json PATH');
-    log('  --out-md PATH');
-    log('  --fixtures-root PATH');
-    log('  --registry PATH');
-    log('  --results-root PATH');
-    log('  --min-pair-margin N  default: 5');
-    log('  --max-pair-solo-wall-ratio N  default: 3');
-    log('  --fail-on-unmeasured');
-    log('');
-    log('Output:', 'green');
-    log('  Prints pair evidence score rows with trigger reasons; --out-md includes a Triggers column');
-    log('');
-    log('Example:', 'green');
-    log('  npx devlyn-cli benchmark frontier --out-md /tmp/devlyn-pair-frontier.md');
-    log('');
-    return;
-  }
-  if (mode === 'audit-headroom') {
-    log('Usage:', 'green');
-    log('  npx devlyn-cli benchmark audit-headroom [options]');
-    log('');
-    log('Options:', 'green');
-    log('  --out-json PATH');
-    log('  --fixtures-root PATH');
-    log('  --registry PATH');
-    log('  --results-root PATH');
-    log('');
-    log('Example:', 'green');
-    log('  npx devlyn-cli benchmark audit-headroom --out-json /tmp/devlyn-headroom-audit.json');
-    log('');
-    return;
-  }
-  if (mode === 'audit') {
-    log('Usage:', 'green');
-    log('  npx devlyn-cli benchmark audit [options]');
-    log('');
-    log('Options:', 'green');
-    log('  --out-dir PATH');
-    log('  --fixtures-root PATH');
-    log('  --registry PATH');
-    log('  --results-root PATH');
-    log('  --min-pair-evidence N  default: 4');
-    log('  --min-pair-margin N  default: 5');
-    log('  --max-pair-solo-wall-ratio N  default: 3');
-    log('  --require-hypothesis-trigger');
-    log('');
-    log('Output:', 'green');
-    log('  Prints frontier score rows plus headroom_rejections=PASS/FAIL, pair_evidence_quality=PASS/FAIL, pair_trigger_reasons=PASS/FAIL, pair_evidence_hypotheses=PASS/FAIL, pair_evidence_hypothesis_triggers=PASS/WARN/FAIL, historical-alias, and hypothesis-trigger gap handoff rows');
-    log('');
-    log('Example:', 'green');
-    log('  npx devlyn-cli benchmark audit --out-dir /tmp/devlyn-benchmark-audit');
-    log('  npx devlyn-cli benchmark audit --require-hypothesis-trigger --out-dir /tmp/devlyn-benchmark-audit-strict');
-    log('');
-    return;
-  }
-  if (mode === 'headroom') {
-    log('Usage:', 'green');
-    log('  npx devlyn-cli benchmark headroom [options] <fixtures...>');
-    log('');
-    log('Options:', 'green');
-    log('  --run-id ID');
-    log('  --bare-max N       default: 60');
-    log('  --solo-max N       default: 80');
-    log('  --min-bare-headroom N  default: 5');
-    log('  --min-solo-headroom N  default: 5');
-    log('  --min-fixtures N   default: 2; use 3 for F16/F23/F25 proof reruns; audit requires 4 passing evidence rows');
-    log('  --allow-rejected-fixtures  active-fixture diagnostics only');
-    log('  --dry-run          validate args/fixtures and print replay command only');
-    log('');
-    log('Example:', 'green');
-    log('  npx devlyn-cli benchmark headroom --min-fixtures 3 F16-cli-quote-tax-rules F23-cli-fulfillment-wave F25-cli-cart-promotion-rules');
-    log('');
-    return;
-  }
-  if (mode === 'pair') {
-    log('Usage:', 'green');
-    log('  npx devlyn-cli benchmark pair [options] <fixtures...>');
-    log('');
-    log('Options:', 'green');
-    log('  --run-id ID');
-    log('  --bare-max N');
-    log('  --solo-max N');
-    log('  --min-bare-headroom N  default: 5');
-    log('  --min-solo-headroom N  default: 5');
-    log('  --min-fixtures N   default: 2; use 3 for F16/F23/F25 proof reruns; audit requires 4 passing evidence rows');
-    log('  --min-pair-margin N  default: 5');
-    log('  --max-pair-solo-wall-ratio N  default: 3');
-    log('  --pair-arm ARM  default: l2_risk_probes; l2_gated is diagnostic');
-    log('  --reuse-calibrated-from RUN_ID');
-    log('  --allow-rejected-fixtures  active-fixture diagnostics only');
-    log('  --dry-run       validate args/fixtures and print replay command only');
-    log('');
-    log('Example:', 'green');
-    log('  npx devlyn-cli benchmark pair --min-fixtures 3 --max-pair-solo-wall-ratio 3 F16-cli-quote-tax-rules F23-cli-fulfillment-wave F25-cli-cart-promotion-rules');
-    log('');
-    return;
-  }
-  showBenchmarkHelp();
-}
-
 // Main
 const args = process.argv.slice(2);
 
@@ -1234,45 +1061,6 @@ switch (command) {
   case 'ls':
     listContents();
     break;
-  case 'benchmark':
-  case 'bench': {
-    const benchmarkScripts = {
-      suite: 'run-suite.sh',
-      recent: 'recent-benchmark-summary.py',
-      frontier: 'pair-candidate-frontier.py',
-      audit: 'audit-pair-evidence.py',
-      'audit-headroom': 'audit-headroom-rejections.py',
-      headroom: 'run-headroom-candidate.sh',
-      pair: 'run-full-pipeline-pair-candidate.sh',
-    };
-    let forwardedArgs = args.slice(1);
-    if (forwardedArgs[0] === '--help' || forwardedArgs[0] === '-h') {
-      showBenchmarkHelp();
-      break;
-    }
-    let benchmarkMode = 'suite';
-    if (forwardedArgs[0] === 'suite' || forwardedArgs[0] === 'recent' || forwardedArgs[0] === 'frontier' || forwardedArgs[0] === 'audit' || forwardedArgs[0] === 'audit-headroom' || forwardedArgs[0] === 'headroom' || forwardedArgs[0] === 'pair') {
-      benchmarkMode = forwardedArgs[0];
-      forwardedArgs = forwardedArgs.slice(1);
-    }
-    if (forwardedArgs[0] === '--help' || forwardedArgs[0] === '-h') {
-      showBenchmarkModeHelp(benchmarkMode);
-      break;
-    }
-    const runnerName = benchmarkScripts[benchmarkMode];
-    const runner = path.join(__dirname, '..', 'benchmark', 'auto-resolve', 'scripts', runnerName);
-    if (!fs.existsSync(runner)) {
-      log('❌ Benchmark suite runner missing — is this a clean devlyn-cli checkout?', 'yellow');
-      log(`   Expected: ${runner}`, 'dim');
-      process.exit(1);
-    }
-    const { spawnSync } = require('child_process');
-    const env = { ...process.env, DEVLYN_BENCHMARK_CLI_SUBCOMMAND: benchmarkMode };
-    const executable = (benchmarkMode === 'recent' || benchmarkMode === 'frontier' || benchmarkMode === 'audit' || benchmarkMode === 'audit-headroom') ? 'python3' : 'bash';
-    const res = spawnSync(executable, [runner, ...forwardedArgs], { stdio: 'inherit', env });
-    process.exit(res.status ?? 1);
-    break;
-  }
   case 'agents': {
     showLogo();
     log('─'.repeat(44), 'dim');

package/bin/instruction-templates.json

@@ -131,6 +131,7 @@
       "3475b34e03725d5e8ea4f2e2fbecc60e1c3ad4a89167e1bb9a26d41640eccbd7",
       "34a09b8d6023ae5b858ec5135cae8efeabd921e3108e63e9627a222b188fe554",
       "350aa6a3140d56a22cb4d5f2503f453ad238384150dc2b5e480fc626db33dfec",
+      "37f003566c80ead28d713e86bd6a877e2e3239302c884911f5144a896c291c17",
       "39af5b94a2734e4fc6c943b604fde4b0d798c4a2a7b1ffb5d053ab7f22488de3",
       "39c3033487667dc0020f9cb6c106afb000177294560be80d67b5020475aec3f6",
       "3c38ab8745dde6391d4dda47d74fe5f8cef410c28325d43dc0dd22e8cf4ab4a7",
@@ -221,6 +222,7 @@
       "9f62c5b2a94e8b5e81528ccbdb6f78d147d14882108568b2cb3b7f85212694c0",
       "a02eb2d8d77553327fb77377dbfedeab059b2e68e0a33c469b9be54a5d55182b",
       "a3a15249968f64085b0cec8c0826688a4c3a1e4cf09b4879054baa086d5cbca1",
+      "a4da26a46ab83815d8ca5d747c68a907ad67dad485adbc6c863b982ee23b307e",
       "a5d6ff585d696254d6727cd1e6d7f8825881f90210eff744265e211ced715ea3",
       "a7ff913df40945a8a16b6632214020a368119ff665953d4a07ec58330b39f101",
       "aafc680a7ac73e100f4a17d2c4070190e68563f78614c39f4c2f58f1e291d9a9",
@@ -255,6 +257,7 @@
       "caaed0b9f9f1e425dcfa45b098c9cde91fc06e50339600585bb7597dd634f636",
       "cb72ba4857899e510a64d8365529e46593e8c69b829d0b5b90e611dd9f1fbdc0",
       "cc8bf51efb9ef8e5fa4134743f2d25434f4fd51595128edcfd2331dbf29d7dd5",
+      "ccacf091bab4102d7156631c6227120a661bb4ec0df514e14a59db44b7bd5dfd",
       "ce1b0d9d0621fd057e684dc18b43156d30f589c72ed20189ea1232ea4f7ce6c6",
       "cebacda3b11db7aa161d96b12edd6f3f790ee2850404cd42ecdf2912ed53de08",
       "cefad7acb7dcd92777f888081aa6585b1418aa581366736fd44db68b1bd66dd0",
@@ -270,6 +273,7 @@
       "ddb200267b43942de3c061b449d08cc3f73441258b74816cfe13da091377b428",
       "dff43f9976c68267baf13f6a460f24af8e76be233c8440ac4684dc664106deae",
       "e237f12ed08800ccc4c6eb4ae57cdd59d61198711f6a3b02dcb668aa08fddb93",
+      "e289dbab99dadb30d8315b32b8243154148c82955f238936bce4c1fe1c3ae525",
       "e2bac40f3ff0b39726f95caaa9b19aa3b21596f165ac7a9dc68ebf42681d58cb",
       "e3ee2da50e961046da3f5b68b0c328570307fb21a93424ae22536b8f216b55a3",
       "e43e4b5b86044fbf7be8d7fd1e3e27cd2089bfc8243e07dc2c3fc034c19794ea",
@@ -400,6 +404,7 @@
       "287fa116a8235756a8b0e28caa951a9ad981667443434e13dbfde3c23f1590e6",
       "28c5b42a58dca1a093576c8c35369c346caaf3963e14a4702ab3330d0d93e4d4",
       "2939f7a22fd12f3ca3fdd0a08fc52535cf25dae38b9b445aceed09dc331fc1de",
+      "2969b7a979fd58d76a2b274270edaf5c08bb8a09e1f25bae34d81a350742ecb2",
       "2989b31e2b03781fb27bdf016db97045772eaaf7459a68728cbbfaa7df916a11",
       "2a23d1450b104b591b0c1358c181bd3ec30c32b8e3a477c342c31272117fdc9d",
       "2a2d82124b77fb1a077bae0a31c849e15ad4b8a3221f5252c99955dd1b835b1b",
@@ -459,6 +464,8 @@
       "43fc337e91354bc672daafc86122b6f15b33c9ca71097a6a056fc55bb0fba1c4",
       "440ddccccc9060a3f5c1a69568e95197b4238cdb4992c4de605d62d6990590e3",
       "4428ea64057045d076c80488245cf7461d45fa7b2e966de5f65292458f96e718",
+      "447d13ec6bc76320d5d305a69c3154a707d7ca0054c170d5d8a18e8c6f7e4001",
+      "45335f760ad0ac3f1fb23ebc9d3ed4e4bf325656dadd4e9cedb9c31c12804361",
       "4549d24fa764d8cd240c7eae696e875378590aa7a27097dfeceb90c7a4bb9291",
       "4609f53455366937df6cbd0f2b4cd19972ebe2db4c89a470fffd44b98583a6f2",
       "462d1647f9e1506d5fc8cd31e2ce9cccb3cba0063193a2145007126ea7e8ed59",
@@ -769,6 +776,7 @@
       "e29fec7d54a144ea4867a3764fccdee6849d58d0f32eacdd56b2c317863fa1c6",
       "e2e0fc585321ee2013cccaed95d47c49fbca71b54c67abad9080d19d70e3c1a5",
       "e3b2f7be15a793f07cbe8e51aa2feeac51d4c67077b94e5e20bad2e14be07fc5",
+      "e468c242cc38c390f9679fca61f355848f58d51570249dfbb65b4f2e26a9dfb1",
       "e4bafa0f553e899b65f05feabf4cbe86d8246c57cbba745eb79e6aa858f40a6b",
       "e5160f4da789914e7f93694a55b9f80b8d06c3b0a126ff60f7627b0b418ae292",
       "e55a62fa1a993f5df1b44d1c55ef27915938d58a1c357604f3dc150f795672df",

package/config/skills/_shared/adapters/grok.md

@@ -67,8 +67,7 @@ tool-use loop (2/3 made zero tool calls and fabricated findings), and a
 
 The judge prompt must name the exact permitted command(s). Derive the prompt
 anchor and allow-list string identically from one source, in this precedence:
-the solo-headroom hypothesis's backticked observable command when present;
-otherwise the backticked commands in the spec's `## Verification` bullets;
+the backticked commands in the spec's `## Verification` bullets;
 otherwise the repo's existing test/CLI runner. If none exists, the probe
 obligation does not arise and a static review is correct for that spec.
 The documented `Bash(<command>)` form matches the exact command or its prefix;

package/config/skills/_shared/archive_run.py

@@ -110,6 +110,9 @@ BOOTSTRAP_TEMP_PATTERNS = tuple(
 PER_RUN_PATTERNS += BOOTSTRAP_TEMP_PATTERNS
 
 SAFE_RUN_ID_RE = re.compile(r"^[A-Za-z0-9_.-]+$")
+valid_final_verdict = runpy.run_path(
+    pathlib.Path(__file__).with_name("terminal-claim-check.py")
+)["valid_final_verdict"]
 
 
 class ArchiveError(Exception):
@@ -173,9 +176,7 @@ def read_run_id(devlyn: pathlib.Path) -> str:
 def is_completed(state: object) -> bool:
     phases = state.get("phases") if isinstance(state, dict) else None
     final = phases.get("final_report") if isinstance(phases, dict) else None
-    if not isinstance(final, dict) or final.get("verdict") not in (
-        "PASS", "PASS_WITH_ISSUES", "NEEDS_WORK", "BLOCKED",
-    ):
+    if not isinstance(final, dict) or not valid_final_verdict(final.get("verdict")):
         return False
     timestamp = final.get("completed_at")
     if not isinstance(timestamp, str) or not re.fullmatch(
@@ -566,6 +567,16 @@ def completion_self_test() -> None:
     print("PASS archive completion: 8 terminal forms; malformed/unfinished/residue retained beyond keep-10")
 
 
+def full_terminal_verdict_completion_self_test() -> None:
+    timestamp = "2026-09-06T00:00:00Z"
+    assert is_completed({"phases": {"final_report": {
+        "verdict": "BLOCKED:repair-budget-exhausted", "completed_at": timestamp,
+    }}})
+    assert not is_completed({"phases": {"final_report": {
+        "verdict": "BLOCKED:", "completed_at": timestamp,
+    }}})
+
+
 def self_test() -> int:
     assert dynamic_judge_role_artifacts(pathlib.Path(".devlyn"), {"phases": {"verify": None}}) == []
     assert dynamic_judge_role_artifacts(pathlib.Path(".devlyn"), {"phases": {}}) == []
@@ -921,6 +932,7 @@ def self_test() -> int:
         else:
             raise AssertionError("NaN archive run_id was accepted")
     completion_self_test()
+    full_terminal_verdict_completion_self_test()
     return 0