catherd-cli 0.1.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Vigen
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,83 @@
+# catherd
+
+```
+ /\_/\  .  catherd
+(=^.^=)/   herds coding agents
+ (")(")
+```
+
+Autopilot builds from your own Claude Code session. Claude plans and verifies, Codex and opencode
+workers write the code, and [Jev](https://typesafe.ai) picks the model and effort for each piece
+of work, climbing a ladder only when a cheaper rung falls short.
+
+- **Your harness, as you set it up.** Every role runs in its vendor's own CLI with your config,
+  hooks, skills and `AGENTS.md`. Isolation is an opt-in toggle per profile and harness, for when
+  you'd rather save the tokens your customizations cost.
+- **Claude roles stay native.** Claude models run as ordinary Claude Code subagents; Codex and
+  opencode run through catherd's MCP server.
+- **Survives restarts.** Workers are detached processes writing straight to disk, so a dropped
+  MCP server never loses a run.
+- **Guard rails for autopilot.** Quota failover to another backend, a preflight check before any
+  worker starts, per-repo knowledge carried between runs, and a run budget.
+
+## Requirements
+
+- [Bun](https://bun.sh) ≥ 1.4
+- [Claude Code](https://claude.com/claude-code) (desktop app or CLI)
+- At least one worker backend: [Codex CLI](https://github.com/openai/codex) and/or
+  [opencode](https://opencode.ai), logged in
+- A TypeSafe API key for Jev, in `TYPESAFE_API_KEY` or `~/.config/typesafe/api_key`
+
+## Install
+
+```sh
+bunx catherd-cli init
+```
+
+The npm package is `catherd-cli`; the command it installs is `catherd`. `init` asks for the Jev key, finds your backends, and builds your first profile. Then add the
+plugin to Claude Code:
+
+```sh
+claude plugin marketplace add 47vigen/catherd
+claude plugin install catherd@catherd
+```
+
+Start a new Claude Code session so the plugin, its MCP server and the agent files load.
+
+## Use
+
+In Claude Code:
+
+- `/catherd <task>` runs a task on autopilot; `/catherd` alone resumes the latest run.
+- `/catherd-setup` tunes your profile in conversation: which models and efforts each role may
+  use, cost or speed, isolation, budget and failover.
+
+In a terminal:
+
+| Command                            | What it does                                                     |
+| ---------------------------------- | ---------------------------------------------------------------- |
+| `bunx catherd-cli`                 | Profile editor: models and efforts per role, with paw checkboxes |
+| `bunx catherd-cli watch`           | Live view of running and recent runs                             |
+| `bunx catherd-cli lock -- <cmd>`   | Runs a heavy command behind the machine-wide semaphore           |
+| `bunx catherd-cli catalog refresh` | Refreshes the models.dev snapshot                                |
+
+`--plain` and `--reduced-motion` work everywhere.
+
+Run data lives in `~/.local/share/catherd/`, config in `~/.config/catherd/` (both follow
+`XDG_*`).
+
+## Develop
+
+```sh
+bun install
+bun test
+bun run typecheck && bun run lint && bun run format:check
+```
+
+Design: [`docs/specs/2026-09-24-catherd-design.md`](docs/specs/2026-09-24-catherd-design.md).
+Releases go through [Changesets](https://github.com/changesets/changesets): add one with
+`bunx changeset`.
+
+## License
+
+MIT

package/catalog/catalog.json

@@ -0,0 +1,112 @@
+{
+  "version": "2026-09-24",
+  "sources": {
+    "repo_code": "DeepSWE-style resolution of real repository issues, per model and effort, from the published results collected in spec §5 on 2026-09-24",
+    "terminal": "Terminal-Bench-style terminal, ops and environment work, per model and effort, from the same collection",
+    "honesty": "100 minus the rate of failing to disclose a broken tool, from the system cards: gpt-6-luna 71.3, gpt-6-sol 95.1, claude-opus-5-5 not published",
+    "secs_per_task": "median seconds per task in the same benchmark runs",
+    "costRank": "catherd's own order of what one run costs a user on a ChatGPT plan and a Claude plan, lowest first; subscription quota counts"
+  },
+  "notes": {
+    "gpt-6-sol": "context 400000 is unverified for the ChatGPT-login id",
+    "gpt-6-luna": "context 400000 is unverified for the ChatGPT-login id",
+    "claude-opus-5-5": "context 1000000 is unverified",
+    "unscored on purpose": "gpt-6-luna#xhigh was dominated; gpt-6-luna#max took 958 s a task, slower than gpt-6-sol#xhigh at an equal score"
+  },
+  "models": [
+    {
+      "id": "gpt-6-luna",
+      "backend": "codex",
+      "efforts": ["low", "medium", "high", "xhigh", "max"],
+      "capabilities": {
+        "toolCall": true,
+        "imageIn": true,
+        "imageOut": true,
+        "reasoning": true,
+        "context": 400000
+      }
+    },
+    {
+      "id": "gpt-6-sol",
+      "backend": "codex",
+      "efforts": ["low", "medium", "high", "xhigh"],
+      "capabilities": {
+        "toolCall": true,
+        "imageIn": true,
+        "imageOut": true,
+        "reasoning": true,
+        "context": 400000
+      }
+    },
+    {
+      "id": "claude-opus-5-5",
+      "backend": "claude",
+      "efforts": ["low", "medium", "high"],
+      "capabilities": {
+        "toolCall": true,
+        "imageIn": true,
+        "imageOut": false,
+        "reasoning": true,
+        "context": 1000000
+      }
+    }
+  ],
+  "entries": [
+    {
+      "rung": "gpt-6-luna#high",
+      "scores": { "repo_code": 59.3, "terminal": 4.5, "honesty": 71.3, "secs_per_task": 563 },
+      "costRank": 1
+    },
+    {
+      "rung": "gpt-6-sol#medium",
+      "scores": { "repo_code": 56.6, "terminal": 18.7, "honesty": 95.1, "secs_per_task": 264 },
+      "costRank": 2
+    },
+    {
+      "rung": "gpt-6-sol#high",
+      "scores": { "repo_code": 65.3, "terminal": 26.3, "honesty": 95.1, "secs_per_task": 391 },
+      "costRank": 3
+    },
+    {
+      "rung": "gpt-6-sol#xhigh",
+      "scores": { "repo_code": 66.6, "terminal": 30.3, "honesty": 95.1, "secs_per_task": 496 },
+      "costRank": 4
+    },
+    { "rung": "claude-opus-5-5#low", "scores": { "terminal": 31.3 }, "costRank": 5 },
+    { "rung": "claude-opus-5-5#medium", "scores": { "terminal": 52.5 }, "costRank": 6 },
+    { "rung": "claude-opus-5-5#high", "scores": { "terminal": 56.6 }, "costRank": 7 }
+  ],
+  "bars": {
+    "repo_code": {
+      "copy": { "repo_code": 55 },
+      "build": { "repo_code": 55 },
+      "logic": { "repo_code": 55, "terminal": 15 },
+      "hard": { "repo_code": 56, "terminal": 18 }
+    },
+    "terminal": {
+      "copy": { "terminal": 15 },
+      "build": { "terminal": 15 },
+      "logic": { "terminal": 15 },
+      "hard": { "terminal": 15 }
+    },
+    "ui": {
+      "copy": { "repo_code": 55 },
+      "build": { "repo_code": 55 },
+      "logic": { "repo_code": 55, "terminal": 15 },
+      "hard": { "repo_code": 56, "terminal": 18 }
+    },
+    "prose": {
+      "copy": { "repo_code": 55 },
+      "build": { "repo_code": 55 },
+      "logic": { "repo_code": 55, "terminal": 15 },
+      "hard": { "repo_code": 56, "terminal": 18 }
+    },
+    "research": {
+      "copy": { "repo_code": 55 },
+      "build": { "repo_code": 55 },
+      "logic": { "repo_code": 55, "terminal": 15 },
+      "hard": { "repo_code": 56, "terminal": 18 }
+    }
+  },
+  "treatLike": {}
+}