devlyn-cli 3.2.1 → 3.3.0

package/benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/pair-plan.sample-pass.json

@@ -1,223 +0,0 @@
-{
-  "accepted_invariants": [
-    {
-      "authority": "expected.json/forbidden_patterns",
-      "id": "forbidden_pattern__silent_catch_returning_a_fallback_value_violates_no_silent_c__bin_cli_js",
-      "operational_check": "variant arm output MUST NOT contain regex pattern \"catch\\\\s*\\\\([^)]*\\\\)\\\\s*\\\\{[^}]*return\\\\s+(\\\\[\\\\]|null|undefined|\\\\{|false|'')\" in files ['bin/cli.js']; rationale: silent catch returning a fallback value — violates no-silent-catches policy",
-      "paraphrase": "variant arm output MUST NOT contain regex pattern \"catch\\\\s*\\\\([^)]*\\\\)\\\\s*\\\\{[^}]*return\\\\s+(\\\\[\\\\]|null|undefined|\\\\{|",
-      "source_refs": [
-        "expected.json:forbidden_patterns[0]"
-      ]
-    },
-    {
-      "authority": "expected.json/forbidden_patterns",
-      "id": "forbidden_pattern__ts_ignore_escape_hatch__bin_cli_js",
-      "operational_check": "variant arm output MUST NOT contain regex pattern '@ts-ignore' in files ['bin/cli.js']; rationale: @ts-ignore escape hatch",
-      "paraphrase": "variant arm output MUST NOT contain regex pattern '@ts-ignore' in files ['bin/cli.js']; rationale: @ts-ignore escape hat",
-      "source_refs": [
-        "expected.json:forbidden_patterns[1]"
-      ]
-    },
-    {
-      "authority": "expected.json/max_deps_added",
-      "id": "max_deps_added__0",
-      "operational_check": "variant arm MUST NOT add more than 0 new npm dependencies (count delta of package.json:dependencies + devDependencies)",
-      "paraphrase": "variant arm MUST NOT add more than 0 new npm dependencies (count delta of package.json:dependencies + devDependencies)",
-      "source_refs": [
-        "expected.json:max_deps_added"
-      ]
-    },
-    {
-      "authority": "expected.json/required_files",
-      "id": "required_file__bin_cli_js",
-      "operational_check": "variant arm output MUST contain file 'bin/cli.js' (created or preserved)",
-      "paraphrase": "variant arm output MUST contain file 'bin/cli.js' (created or preserved)",
-      "source_refs": [
-        "expected.json:required_files[bin/cli.js]"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "scope-tier-a:lockfile-deletion",
-      "operational_check": "variant arm MUST NOT delete a scaffold-present lockfile",
-      "paraphrase": "variant arm MUST NOT delete a scaffold-present lockfile",
-      "source_refs": [
-        "oracle-scope-tier-a.py"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "scope-tier-a:tier-a-violation",
-      "operational_check": "variant arm MUST NOT add or modify paths matching: docs/roadmap/** | docs/VISION.md | docs/ROADMAP.md | .github/** | node_modules/** | **/node_modules/** | test-results/** | coverage/** | .nyc_output/** | basename suffix .log | basename prefix .env or secrets.",
-      "paraphrase": "variant arm MUST NOT add or modify paths matching: docs/roadmap/** | docs/VISION.md | docs/ROADMAP.md | .github/** | nod",
-      "source_refs": [
-        "oracle-scope-tier-a.py"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "scope-tier-b:scope-unmatched",
-      "operational_check": "every variant-touched file MUST be either inside spec_output_files (Tier C) OR reachable from a Tier C seed via static JS/TS imports OR matched by expected.json:tier_a_waivers",
-      "paraphrase": "every variant-touched file MUST be either inside spec_output_files (Tier C) OR reachable from a Tier C seed via static J",
-      "source_refs": [
-        "oracle-scope-tier-b.py"
-      ]
-    },
-    {
-      "authority": "expected.json/spec_output_files",
-      "id": "spec_output_file__bin_cli_js",
-      "operational_check": "variant-touched files MUST be inside (or reachable via static imports from) the spec_output_files set; 'bin/cli.js' is one Tier C seed",
-      "paraphrase": "variant-touched files MUST be inside (or reachable via static imports from) the spec_output_files set; 'bin/cli.js' is o",
-      "source_refs": [
-        "expected.json:spec_output_files[bin/cli.js]"
-      ]
-    },
-    {
-      "authority": "expected.json/spec_output_files",
-      "id": "spec_output_file__tests_cli_test_js",
-      "operational_check": "variant-touched files MUST be inside (or reachable via static imports from) the spec_output_files set; 'tests/cli.test.js' is one Tier C seed",
-      "paraphrase": "variant-touched files MUST be inside (or reachable via static imports from) the spec_output_files set; 'tests/cli.test.j",
-      "source_refs": [
-        "expected.json:spec_output_files[tests/cli.test.js]"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "test-fidelity:assertion-regression",
-      "operational_check": "effective assertion count MUST NOT drop and skipped-test count MUST NOT rise; vacuous expect.assertions(0) is treated as a real regression",
-      "paraphrase": "effective assertion count MUST NOT drop and skipped-test count MUST NOT rise; vacuous expect.assertions(0) is treated as",
-      "source_refs": [
-        "oracle-test-fidelity.py"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "test-fidelity:mock-swap",
-      "operational_check": "post-arm test file MUST NOT swap REAL_PATTERNS hits for MOCK_PATTERNS hits (jest/vi/sinon, nock/msw, app.handle/inject/callback, hand-rolled IncomingMessage/ServerResponse, etc.); a drop in real_calls combined with a rise in mock_calls is a mock-swap flag",
-      "paraphrase": "post-arm test file MUST NOT swap REAL_PATTERNS hits for MOCK_PATTERNS hits (jest/vi/sinon, nock/msw, app.handle/inject/c",
-      "source_refs": [
-        "oracle-test-fidelity.py"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "test-fidelity:test-file-deleted",
-      "operational_check": "no scaffold-present test file may be deleted by the variant arm; deletion of an existing tests/*.test.* / *.spec.* / *.e2e.* file is a flag-severity finding",
-      "paraphrase": "no scaffold-present test file may be deleted by the variant arm; deletion of an existing tests/*.test.* / *.spec.* / *.e",
-      "source_refs": [
-        "oracle-test-fidelity.py"
-      ]
-    },
-    {
-      "authority": "metadata/oracle-allowlist",
-      "id": "test-fidelity:test-file-renamed",
-      "operational_check": "rename of a scaffold-present test file is warn-severity (content fidelity not verified across renames in step 1)",
-      "paraphrase": "rename of a scaffold-present test file is warn-severity (content fidelity not verified across renames in step 1)",
-      "source_refs": [
-        "oracle-test-fidelity.py"
-      ]
-    },
-    {
-      "authority": "expected.json/verification_commands",
-      "id": "verification__3f35982a",
-      "operational_check": "running `node bin/cli.js doctor` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doctor:']; stdout MUST NOT contain any of ['undefined', 'Error:']",
-      "paraphrase": "running `node bin/cli.js doctor` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doctor:'];",
-      "source_refs": [
-        "expected.json:verification_commands[0]"
-      ]
-    },
-    {
-      "authority": "expected.json/verification_commands",
-      "id": "verification__460fce04",
-      "operational_check": "running `HOME=/nonexistent node bin/cli.js doctor` in the post-arm work dir MUST exit with code 1; stdout MUST contain all of ['/nonexistent']; stdout MUST NOT contain any of []",
-      "paraphrase": "running `HOME=/nonexistent node bin/cli.js doctor` in the post-arm work dir MUST exit with code 1; stdout MUST contain a",
-      "source_refs": [
-        "expected.json:verification_commands[1]"
-      ]
-    },
-    {
-      "authority": "expected.json/verification_commands",
-      "id": "verification__973e287e",
-      "operational_check": "running `python3 -c \"import subprocess; r = subprocess.run(['node', 'bin/cli.js', 'doctor'], capture_output=True); n = r.stdout.count(b'\\x1b['); print(n); exit(0 if n == 0 else 1)\"` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['0']; stdout MUST NOT contain any of []",
-      "paraphrase": "running `python3 -c \"import subprocess; r = subprocess.run(['node', 'bin/cli.js', 'doctor'], capture_output=True); n = r",
-      "source_refs": [
-        "expected.json:verification_commands[2]"
-      ]
-    },
-    {
-      "authority": "expected.json/verification_commands",
-      "id": "verification__d6253a97",
-      "operational_check": "running `node bin/cli.js doctor --help` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doctor']; stdout MUST NOT contain any of []",
-      "paraphrase": "running `node bin/cli.js doctor --help` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doc",
-      "source_refs": [
-        "expected.json:verification_commands[3]"
-      ]
-    },
-    {
-      "authority": "expected.json/verification_commands",
-      "id": "verification__e0f149e4",
-      "operational_check": "running `node bin/cli.js --help` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doctor']; stdout MUST NOT contain any of []",
-      "paraphrase": "running `node bin/cli.js --help` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doctor']; ",
-      "source_refs": [
-        "expected.json:verification_commands[4]"
-      ]
-    },
-    {
-      "authority": "expected.json/verification_commands",
-      "id": "verification__fdbcd321",
-      "operational_check": "running `node bin/cli.js doctor --verbose` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['doctor:']; stdout MUST NOT contain any of ['Error:']",
-      "paraphrase": "running `node bin/cli.js doctor --verbose` in the post-arm work dir MUST exit with code 0; stdout MUST contain all of ['",
-      "source_refs": [
-        "expected.json:verification_commands[5]"
-      ]
-    }
-  ],
-  "authority_order": [
-    "spec.md",
-    "expected.json/rubric",
-    "phase prompt",
-    "model preference"
-  ],
-  "escalated_to_user": [],
-  "fixture_id": "F2-cli-medium-subcommand",
-  "model_stamps": {
-    "claude": {
-      "blocked_ids": [],
-      "model": "claude-opus-4-7",
-      "signed_plan_sha256": "7329b8955a94ac680a7a58d09bfc53ce4de17609495b4bf4658cf2d3a43dacd5",
-      "status": "sign",
-      "timestamp": "2026-04-29T18:30:00Z"
-    },
-    "codex": {
-      "blocked_ids": [],
-      "model": "gpt-5.5",
-      "signed_plan_sha256": "7329b8955a94ac680a7a58d09bfc53ce4de17609495b4bf4658cf2d3a43dacd5",
-      "status": "sign",
-      "timestamp": "2026-04-29T18:31:00Z"
-    }
-  },
-  "plan_status": "final",
-  "planning_mode": "pair",
-  "rejected_alternatives": [],
-  "rounds": [
-    {
-      "claude_draft_sha256": "0000000000000000000000000000000000000000000000000000000000000000",
-      "codex_draft_sha256": "1111111111111111111111111111111111111111111111111111111111111111",
-      "merged_sha256": "2222222222222222222222222222222222222222222222222222222222222222",
-      "note": "sample-pass synthetic round (test fixture)",
-      "round": 1
-    }
-  ],
-  "schema_version": "1",
-  "source": {
-    "canonical_id_registry_path": "benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/expected-pair-plan-registry.json",
-    "canonical_id_registry_sha256": "98ac16e4536ea3ef2e51d3c728982c014211c193a742cea74f1331e4fbba76be",
-    "expected_path": "benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/expected.json",
-    "expected_sha256": "ddef8feba49f20b6957e37840bc6a03e78e554776e380d81ad6390944c72fcab",
-    "rubric_path": "benchmark/auto-resolve/RUBRIC.md",
-    "rubric_sha256": "5b5b709a0b57f7e6f4fbc072af91e1edbc8d7910ae16b9b7be7170616aeaa9af",
-    "spec_path": "benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/spec.md",
-    "spec_sha256": "9b0949c2afd4a522de2bdbbf267d93907fd908bf0f1d0dc5e111ee30ba875bb7"
-  },
-  "unresolved": []
-}

package/benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/setup.sh

@@ -1,5 +0,0 @@
-#!/usr/bin/env bash
-# F2 setup — no base-repo modifications needed. The task starts from the
-# stock test-repo baseline (hello/version CLI) and adds a new subcommand.
-set -e
-exit 0

package/benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/spec.md

@@ -1,55 +0,0 @@
----
-id: "F2-cli-medium-subcommand"
-title: "Add `doctor` subcommand to bench-test-repo CLI"
-status: planned
-complexity: medium
-depends-on: []
----
-
-# F2 Add `doctor` subcommand
-
-## Context
-
-`bench-test-repo` users need a one-command way to diagnose their local
-environment — node version, Claude Code install, plugins, skills — without
-digging through the filesystem. A `doctor` subcommand lands that capability
-inside the CLI itself.
-
-## Requirements
-
-- [ ] `node bin/cli.js doctor` produces a status report and exits 0 on a clean machine.
-- [ ] Node version check — requires `process.version >= v18.0.0`, emits a status line, marks FAIL if below.
-- [ ] `$HOME/.claude/` check — exists as directory AND is writable. Missing → FAIL. Exists but not writable (EACCES) → FAIL with a distinct "permission" message.
-- [ ] Installed plugins scan — read subdirectories of `$HOME/.claude/plugins/cache/` and print a summary line with the count; `--verbose` lists names.
-- [ ] Installed skills scan — count files matching `$HOME/.claude/skills/**/SKILL.md`; print count; `--verbose` lists relative paths.
-- [ ] Colored output with `[OK]` (green), `[WARN]` (yellow), `[FAIL]` (red) via ANSI escape codes **only when `process.stdout.isTTY` is true** — piped output must contain no `\x1b[` sequences.
-- [ ] Summary line: `doctor: <N> ok, <M> warn, <K> fail`.
-- [ ] Exit code: `0` if zero fails, `1` otherwise.
-- [ ] `--verbose` flag expands details for plugins/skills scans.
-- [ ] `node bin/cli.js doctor --help` prints a short help block and exits 0.
-- [ ] `node bin/cli.js --help` lists `doctor` as an available subcommand.
-- [ ] `HOME=/nonexistent node bin/cli.js doctor` prints a FAIL line clearly referencing the missing `/nonexistent/.claude` and exits 1.
-
-## Constraints
-
-- **Zero new npm dependencies.** Use only Node.js built-ins (`fs`, `path`, `os`, `process`).
-- **No silent error catches.** Do not wrap operations in `try { … } catch { return fallbackValue }`. All errors visible to the user with actionable messages.
-- **HOME guard.** If `process.env.HOME` is undefined or empty, emit a clear FAIL line ("HOME environment variable is not set") and exit 1.
-- **EACCES handling.** If `readdirSync` fails with EACCES, emit a permission-specific message quoting the offending path. Do not silently return an empty list.
-
-
-## Out of Scope
-
-- Auto-repair (report only; do not offer to fix detected problems).
-- Checking remote/registry state (npm, GitHub).
-- Any feature requiring a new npm dependency.
-
-## Verification
-
-- `node bin/cli.js doctor` exits 0 on a machine with `~/.claude` present.
-- `HOME=/nonexistent node bin/cli.js doctor` prints a FAIL line referencing `/nonexistent/.claude` and exits 1.
-- `node bin/cli.js doctor | cat` contains no `\x1b[` sequences.
-- `node bin/cli.js doctor --help` prints help, exits 0.
-- `node bin/cli.js --help` mentions `doctor`.
-- `git diff -- package.json` is empty.
-- `node bin/cli.js doctor --verbose` lists plugins and skills.

package/benchmark/auto-resolve/fixtures/F2-cli-medium-subcommand/task.txt

@@ -1,14 +0,0 @@
-Add a `doctor` subcommand to bench-test-repo's CLI (`bin/cli.js`) that diagnoses the local environment. When a user runs `node bin/cli.js doctor`, it should check:
-
-1. Node version is >= v18.0.0
-2. $HOME/.claude/ exists and is writable
-3. How many plugins are installed under $HOME/.claude/plugins/cache/
-4. How many skills are installed (SKILL.md files under $HOME/.claude/skills/)
-
-Each check produces a status line with a tag — `[OK]`, `[WARN]`, or `[FAIL]`. The tags should be colored (green/yellow/red) when output is a TTY, but plain text when piped. End with a summary line like `doctor: 3 ok, 1 warn, 0 fail` and exit 0 if no fails, 1 otherwise.
-
-Add a `--verbose` flag that lists the individual plugins and skill paths. Add `--help` for the subcommand. Make sure `node bin/cli.js --help` mentions `doctor`.
-
-Keep it to Node.js built-ins only (fs, path, os, process) — no new npm dependencies. Handle errors explicitly: if HOME is unset, fail cleanly with a helpful message; if a directory can't be read due to permissions (EACCES), say so with the path.
-
-Running `HOME=/nonexistent node bin/cli.js doctor` should exit 1 and mention the missing `/nonexistent/.claude`.

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/NOTES.md

@@ -1,39 +0,0 @@
-# F21 CLI scheduler priority
-
-## Failure mode
-
-This fixture detects implementations that pass simple scheduling tests while
-missing interaction rules: global priority ordering, earliest-fit placement,
-blocked-interval exclusion, half-open time boundaries, and deterministic output
-ordering.
-
-## Pipeline phase target
-
-PLAN must preserve the ordering and interval invariants. IMPLEMENT must build a
-small scheduling engine without adding dependencies. VERIFY should catch
-counterexamples where local request order or naive overlap checks produce a
-plausible but wrong schedule.
-
-## Why existing fixtures do not cover it
-
-F16 covers checkout arithmetic. F10/F11/F12/F15 cover server behavior. None
-exercise a CLI algorithm where the correct result depends on sorting,
-interval arithmetic, and output ordering at once.
-
-## Measurement status
-
-Pair evidence passed in `20260511-f21-current-riskprobes-v1`: bare `33`,
-solo_claude `66`, pair `99`, margin `+33`, wall `1.47x`,
-arm `l2_risk_probes`, verdict `pair_evidence_passed`.
-
-## Solo-headroom hypothesis
-
-A capable solo_claude baseline is expected to miss global priority ordering
-combined with blocked-interval earliest-fit placement; observable command
-`node "$BENCH_FIXTURE_DIR/verifiers/priority-blocked.js"` exposes the miss.
-
-## Retirement
-
-Retire or replace when both bare and solo consistently exceed the headroom
-thresholds, or if future fixtures cover the same interval-scheduling failure
-mode with better discrimination.

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/expected.json

@@ -1,62 +0,0 @@
-{
-  "verification_commands": [
-    {
-      "cmd": "node --test tests/cli.test.js",
-      "exit_code": 0,
-      "stdout_contains": [],
-      "stdout_not_contains": ["not ok "]
-    },
-    {
-      "cmd": "node \"$BENCH_FIXTURE_DIR/verifiers/priority-blocked.js\"",
-      "exit_code": 0,
-      "stdout_contains": ["\"ok\":true"],
-      "stdout_not_contains": [],
-      "contract_refs": [
-        "Process requests globally by `priority` descending, then `submitted_at` ascending, then `id` ascending.",
-        "For each request, place it on the requested resource at the earliest start minute that is greater than or equal to the request's requested `start`, fits wholly inside one availability window, does not overlap any blocked interval, and does not overlap any already accepted request on that resource.",
-        "A request may end exactly at a window end, but any one-minute overlap with a blocked interval is rejected or moved later."
-      ]
-    },
-    {
-      "cmd": "node \"$BENCH_FIXTURE_DIR/verifiers/error-order.js\"",
-      "exit_code": 0,
-      "stdout_contains": ["\"ok\":true"],
-      "stdout_not_contains": [],
-      "contract_refs": [
-        "If the request references an unknown resource, reject with `{ \"id\": string, \"reason\": \"unknown_resource\" }`.",
-        "`rejected` is ordered in the original request input order.",
-        "Duplicate request ids are invalid input: exit `2`, one JSON error to stderr, no stdout."
-      ]
-    }
-  ],
-  "forbidden_patterns": [
-    {
-      "pattern": "catch\\s*\\([^)]*\\)\\s*\\{[^}]*return\\s+(null|undefined|''|\\{\\})",
-      "description": "silent catch returning fallback in scheduler path",
-      "files": ["bin/cli.js", "tests/cli.test.js"],
-      "severity": "disqualifier"
-    },
-    {
-      "pattern": "catch\\s*\\([^)]*\\)\\s*\\{\\s*\\}",
-      "description": "empty catch block",
-      "files": ["bin/cli.js", "tests/cli.test.js"],
-      "severity": "disqualifier"
-    },
-    {
-      "pattern": "catch\\s*(?:\\([^)]*\\))?\\s*\\{[^}]*ignore[^}]*\\}",
-      "description": "silent catch-ignore cleanup",
-      "files": ["bin/cli.js", "tests/cli.test.js"],
-      "severity": "disqualifier"
-    },
-    {
-      "pattern": "/\\*\\s*eslint-disable",
-      "description": "eslint-disable without scoped justification",
-      "files": ["bin/cli.js"],
-      "severity": "disqualifier"
-    }
-  ],
-  "required_files": ["bin/cli.js", "tests/cli.test.js"],
-  "forbidden_files": [],
-  "max_deps_added": 0,
-  "spec_output_files": ["bin/cli.js", "tests/cli.test.js"]
-}

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/metadata.json

@@ -1,10 +0,0 @@
-{
-  "id": "F21-cli-scheduler-priority",
-  "category": "high-risk",
-  "difficulty": "high",
-  "timeout_seconds": 1500,
-  "required_tools": ["node"],
-  "browser": false,
-  "deps_change_expected": false,
-  "intent": "Add a bench-cli schedule command that assigns appointment requests to resource availability windows using priority, submitted order, blocked intervals, and earliest-fit placement while producing exact JSON accept/reject results."
-}

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/setup.sh

@@ -1,2 +0,0 @@
-#!/usr/bin/env bash
-set -e

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/spec.md

@@ -1,67 +0,0 @@
----
-id: "F21-cli-scheduler-priority"
-title: "Priority appointment scheduler"
-status: planned
-complexity: high
-depends-on: []
----
-
-# F21 Priority appointment scheduler
-
-## Context
-
-Add a `bench-cli schedule --input <path>` command that assigns appointment
-requests to resource availability windows using priority, submitted order,
-blocked intervals, and earliest-fit placement while producing exact JSON
-accept/reject results.
-
-The scheduler is used by downstream automation, so output shape, ordering, and
-failure reasons must be deterministic.
-
-## Requirements
-
-- [ ] `bench-cli schedule --input <path>` reads JSON shaped as `{ "resources": Array<Resource>, "requests": Array<Request> }`.
-- [ ] Each resource has `{ "id": string, "windows": [{ "start": "HH:MM", "end": "HH:MM" }], "blocked": [{ "start": "HH:MM", "end": "HH:MM" }] }`. `blocked` may be empty.
-- [ ] Each request has `{ "id": string, "resource": string, "start": "HH:MM", "duration_min": number, "priority": number, "submitted_at": string }`.
-- [ ] Times are same-day 24-hour clock minutes. A range is half-open: `[start, end)`. A request ending exactly at a window end is allowed; overlapping a blocked range by one minute is not allowed.
-- [ ] Process requests globally by `priority` descending, then `submitted_at` ascending, then `id` ascending.
-- [ ] For each request, place it on the requested resource at the earliest start minute that is greater than or equal to the request's requested `start`, fits wholly inside one availability window, does not overlap any blocked interval, and does not overlap any already accepted request on that resource.
-- [ ] Do not move a request to a different resource.
-- [ ] If no placement exists, reject with `{ "id": string, "reason": "no_slot" }`.
-- [ ] If the request references an unknown resource, reject with `{ "id": string, "reason": "unknown_resource" }`.
-- [ ] Invalid top-level shape, invalid time strings, non-positive or non-integer `duration_min`, or duplicate request ids exits `2`, writes exactly one JSON error object to stderr, and writes nothing to stdout.
-- [ ] On success, write exactly one JSON object to stdout and no stderr. Keys: `scheduled`, `rejected`.
-- [ ] `scheduled` is ordered by actual scheduled start time ascending, then resource id ascending, then request id ascending. Each row has keys `id`, `resource`, `start`, `end`.
-- [ ] `rejected` is ordered in the original request input order. Each row has keys `id`, `reason`.
-- [ ] `tests/cli.test.js` is updated. Existing tests still pass AND at least two scheduler tests cover one success case and one rejection case.
-
-## Constraints
-
-- **No new npm dependencies.**
-- **No silent catches in implementation or tests.** Invalid input and file-read failures must surface as JSON errors with exit `2`; test cleanup should use explicit safe primitives such as `fs.rmSync(path, { force: true })`, not `catch { /* ignore */ }`.
-- **No mutation of the input file.**
-- **No extra stdout/stderr text** on the success path; downstream tooling parses stdout as JSON.
-- **Touch only `bin/cli.js` and `tests/cli.test.js`.**
-
-## Out of Scope
-
-- Multiple-day scheduling.
-- Time zones.
-- Recurring appointments.
-- Persisting scheduled results.
-- Touching `server/`, `web/`, or `tests/server.test.js`.
-
-## Verification
-
-- `node --test tests/cli.test.js` exits 0.
-- A higher-priority later-submitted request can take the first slot, forcing a lower-priority earlier-submitted request to the next non-overlapping slot.
-- A request may end exactly at a window end, but any one-minute overlap with a blocked interval is rejected or moved later.
-- Unknown resources are reported in `rejected` without aborting the whole run.
-- Duplicate request ids are invalid input: exit `2`, one JSON error to stderr, no stdout.
-- `git diff --stat` shows only `bin/cli.js` and `tests/cli.test.js` touched.
-
-## Solo-headroom hypothesis
-
-A capable solo_claude baseline is expected to miss global priority ordering
-across resources while preserving blocked-interval earliest-fit placement;
-observable command `node "$BENCH_FIXTURE_DIR/verifiers/priority-blocked.js"` exposes the miss.

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/task.txt

@@ -1,7 +0,0 @@
-Add a `schedule` command to `bench-cli` so users can run `bench-cli schedule --input <path>` with a JSON file containing resources and appointment requests. It should assign requests to resource availability windows using priority, submitted order, blocked intervals, and earliest-fit placement, then print one exact JSON result with `scheduled` and `rejected` arrays.
-
-The input has resources with `id`, `windows`, and `blocked`, and requests with `id`, `resource`, `start`, `duration_min`, `priority`, and `submitted_at`. Times are same-day `HH:MM` values. Process requests by priority descending, then submitted_at ascending, then id ascending. A placement must stay on the requested resource, start no earlier than the requested start, fit inside one window, avoid blocked intervals, and avoid already accepted requests.
-
-On success, stdout must be exactly parseable JSON and stderr must be empty. `scheduled` rows must be ordered by actual scheduled start time, then resource id, then request id. `rejected` rows must stay in original request input order. Invalid input, duplicate request ids, invalid times, bad durations, or file-read failures should exit `2`, print exactly one JSON error object to stderr, and print nothing to stdout.
-
-Update `tests/cli.test.js` so existing tests still pass and add at least two scheduler tests: one success case and one rejection case. Do not use silent catches in implementation or test cleanup; use explicit safe primitives such as `fs.rmSync(path, { force: true })` instead of `catch { /* ignore */ }`. No new npm dependencies. Only touch `bin/cli.js` and `tests/cli.test.js`.

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/verifiers/error-order.js

@@ -1,55 +0,0 @@
-'use strict';
-const assert = require('node:assert');
-const { execFileSync, spawnSync } = require('node:child_process');
-const fs = require('node:fs');
-const os = require('node:os');
-const path = require('node:path');
-
-const work = process.env.BENCH_WORKDIR || process.cwd();
-const cli = path.join(work, 'bin', 'cli.js');
-const tmp = fs.mkdtempSync(path.join(os.tmpdir(), 'f21-errors-'));
-
-const unknownInput = path.join(tmp, 'unknown.json');
-fs.writeFileSync(unknownInput, JSON.stringify({
-  resources: [
-    { id: 'r1', windows: [{ start: '10:00', end: '10:30' }], blocked: [] }
-  ],
-  requests: [
-    { id: 'unknown-first', resource: 'missing', start: '10:00', duration_min: 5, priority: 9, submitted_at: '2026-01-01T10:00:00Z' },
-    { id: 'too-long', resource: 'r1', start: '10:00', duration_min: 45, priority: 8, submitted_at: '2026-01-01T10:01:00Z' },
-    { id: 'ok', resource: 'r1', start: '10:00', duration_min: 30, priority: 7, submitted_at: '2026-01-01T10:02:00Z' }
-  ]
-}), 'utf8');
-
-const out = execFileSync('node', [cli, 'schedule', '--input', unknownInput], {
-  cwd: work,
-  encoding: 'utf8',
-  stdio: ['ignore', 'pipe', 'pipe']
-});
-const parsed = JSON.parse(out);
-assert.deepStrictEqual(parsed.scheduled, [
-  { id: 'ok', resource: 'r1', start: '10:00', end: '10:30' }
-]);
-assert.deepStrictEqual(parsed.rejected, [
-  { id: 'unknown-first', reason: 'unknown_resource' },
-  { id: 'too-long', reason: 'no_slot' }
-]);
-
-const dupInput = path.join(tmp, 'dup.json');
-fs.writeFileSync(dupInput, JSON.stringify({
-  resources: [
-    { id: 'r1', windows: [{ start: '10:00', end: '11:00' }], blocked: [] }
-  ],
-  requests: [
-    { id: 'dup', resource: 'r1', start: '10:00', duration_min: 10, priority: 1, submitted_at: '2026-01-01T10:00:00Z' },
-    { id: 'dup', resource: 'r1', start: '10:10', duration_min: 10, priority: 1, submitted_at: '2026-01-01T10:01:00Z' }
-  ]
-}), 'utf8');
-const dup = spawnSync('node', [cli, 'schedule', '--input', dupInput], {
-  cwd: work,
-  encoding: 'utf8'
-});
-assert.strictEqual(dup.status, 2);
-assert.strictEqual(dup.stdout, '');
-assert.doesNotThrow(() => JSON.parse(dup.stderr));
-console.log(JSON.stringify({ ok: true }));

package/benchmark/auto-resolve/fixtures/F21-cli-scheduler-priority/verifiers/priority-blocked.js

@@ -1,48 +0,0 @@
-'use strict';
-const assert = require('node:assert');
-const { execFileSync } = require('node:child_process');
-const fs = require('node:fs');
-const os = require('node:os');
-const path = require('node:path');
-
-const work = process.env.BENCH_WORKDIR || process.cwd();
-const cli = path.join(work, 'bin', 'cli.js');
-const tmp = fs.mkdtempSync(path.join(os.tmpdir(), 'f21-schedule-'));
-const input = path.join(tmp, 'input.json');
-
-fs.writeFileSync(input, JSON.stringify({
-  resources: [
-    {
-      id: 'room-a',
-      windows: [{ start: '09:00', end: '10:00' }],
-      blocked: [{ start: '09:30', end: '09:40' }]
-    },
-    {
-      id: 'room-b',
-      windows: [{ start: '09:00', end: '09:45' }],
-      blocked: []
-    }
-  ],
-  requests: [
-    { id: 'low-first', resource: 'room-a', start: '09:00', duration_min: 30, priority: 1, submitted_at: '2026-01-01T09:00:00Z' },
-    { id: 'high-second', resource: 'room-a', start: '09:00', duration_min: 30, priority: 10, submitted_at: '2026-01-01T09:05:00Z' },
-    { id: 'edge-ok', resource: 'room-b', start: '09:15', duration_min: 30, priority: 5, submitted_at: '2026-01-01T09:01:00Z' },
-    { id: 'blocked-one-minute', resource: 'room-a', start: '09:29', duration_min: 2, priority: 4, submitted_at: '2026-01-01T09:02:00Z' }
-  ]
-}), 'utf8');
-
-const stdout = execFileSync('node', [cli, 'schedule', '--input', input], {
-  cwd: work,
-  encoding: 'utf8',
-  stdio: ['ignore', 'pipe', 'pipe']
-});
-const parsed = JSON.parse(stdout);
-assert.deepStrictEqual(parsed.scheduled, [
-  { id: 'high-second', resource: 'room-a', start: '09:00', end: '09:30' },
-  { id: 'edge-ok', resource: 'room-b', start: '09:15', end: '09:45' },
-  { id: 'blocked-one-minute', resource: 'room-a', start: '09:40', end: '09:42' }
-]);
-assert.deepStrictEqual(parsed.rejected, [
-  { id: 'low-first', reason: 'no_slot' }
-]);
-console.log(JSON.stringify({ ok: true }));

package/benchmark/auto-resolve/fixtures/F22-cli-ledger-close/NOTES.md

@@ -1,35 +0,0 @@
-# F22 CLI ledger close
-
-## Failure mode
-
-This fixture detects finance-style reconciliation mistakes: applying duplicate
-transactions twice, silently accepting conflicting duplicate ids, validating
-only while mutating balances, rejecting negative balances that are explicitly
-allowed, or producing nondeterministic account ordering.
-
-## Pipeline phase target
-
-PLAN must separate validation, idempotency, chronological application, and
-output formatting. IMPLEMENT must keep cents as integers and avoid fallback
-error handling. VERIFY should catch duplicate-id counterexamples and negative
-balance behavior.
-
-## Why existing fixtures do not cover it
-
-F16 covers order quote arithmetic, but not ledger idempotency or full-input
-validation before mutation. F21 covers interval scheduling. Server fixtures
-cover API behavior rather than CLI reconciliation.
-
-## Measurement status
-
-Headroom runs reject F22 as full-pipeline pair-lift evidence. In
-`20260507-f21-f22-full-pipeline-pair`, F22 scored bare 91 / solo_claude 98 and
-failed the headroom gate. In `20260508-f22-exact-error-headroom`, F22 scored
-bare 94 / solo_claude 98 after the exact-error fixture revision. Keep it as a ledger
-reconciliation control, not as counted `solo < pair` evidence.
-
-## Retirement
-
-Retire or replace if both bare and solo consistently score above the headroom
-thresholds, or if a future ledger fixture captures the same duplicate-id and
-validation-before-mutation risks with stronger discrimination.