Muse Code vs Sculptor

Muse Code 20

Harness reading of the published rubric. No payment protocol (0). Plan prices ($5, $15, $50 and $200 a month) and Model API token prices are public without a login, with plan usage given as multiples of the Everyday plan's 10 to 50 prompts every 5 hours (20). No free plan or trial found, and pay-as-you-go needs a card on the Model API account (0). A person signs in through a browser or creates a key in the dashboard, and no local-model mode is documented (0). 0 + 20 + 0 + 0 = 20.

Sculptor 60

Free self-hosted software with nothing to buy, so 0 for a protocol and 20 each for public pricing, a free tier and onboarding with no sign-up, 60.

Muse Code 84

Framework reading. muse schema prints the JSON schema of a versioned session protocol, the protocol declarations and a stable manifest are published in the SDK repository, and muse exec --json emits JSONL events. No JSON Schema for settings.json was found (22 of 25). llms.txt at dev.meta.ai/llms.txt with a Markdown copy of each page (10). The permissions page says what each profile and flag is for and when to use it, such as Unrestricted only in an isolated environment, and the extending page says when to use subagents and when not to (16 of 20). Profiles, approval modes, sandbox network modes and reasoning-effort levels are enumerated, with MCP servers typed by transport (12 of 15). Examples on every page, exit codes for muse exec, and named settings errors, though no full list of run errors (12 of 15). A changelog for every version to 1.4.13, with no dates (12 of 15). 22 + 10 + 16 + 12 + 12 + 12 = 84.

Sculptor 45

Harness reading. A README, specs and agent docs in the repository; no API reference (45).

Muse Code 79

Framework reading, adapted to a harness driven by a pipeline. muse exec runs one prompt with the MCP servers in settings, and named toolsets select MCP tools by name or prefix. We found no deferred tool loading in Muse Code's own docs (18 of 25). --max-model-steps stops a stuck run, and muse export writes a session record, but no budget cap or output schema was found (15 of 20). Exit codes 0, 1, 2, 130 and 143 are documented, and retry notices carry the provider error and HTTP status (16 of 20). --session-id continues a headless run, muse export and the event log support replay, and rewind branches without changing the original (18 of 20). TypeScript and Python SDKs. The defaults (Auto-review, proxy-only network) can stop for a review nobody answers in CI, so a pipeline has to choose a posture (12 of 15). 18 + 15 + 16 + 18 + 12 = 79.

Sculptor 50

Harness reading. Workspaces as isolated copies, review and merge back, and built-in skills (50).

Muse Code 67

Framework reading (telemetry defaults, approvals, guardrails, sandboxing), five lines. A browser session or a Model API key, cleared by muse logout. Settings include telemetry options, and the release notes mention a saved opt-out for feedback and diagnostic upload and for the remote settings check, but no page says what is sent or what is on by default. Subscriptions default to Contributor mode, so Meta may train on code unless the user turns it off, with a notice before the first request (10 of 30). Ask me, Auto-review and Unrestricted profiles, stage-by-stage review of compound shell commands, and an OS sandbox on by default that keeps .git, .muse and .agents read-only and refuses to run when it can't confirm containment. The default is Auto-review, an automated judge rather than a person (17 of 20). Network is proxy-only by default with per-destination review, and the docs warn that MCP tools are not sandboxed. The 1.4.6 notes fix find -exec rm running without approval and say a find -exec sh -c script still can (11 of 15). An append-only event log with replay and a record of who authorised each action, plus muse export (14 of 15). Meta's bug bounty, named in the SDK's SECURITY.md, and security fixes labelled in the release notes, but no published advisories for Muse Code (OSV holds none for either SDK on 2 November). The 12 October reading of www.facebook.com's security.txt is struck (that host's robots.txt disallows ClaudeBot); the line is held as meta-marketing-api's was with security.txt unread (15 of 20). 10 + 17 + 11 + 14 + 15 = 67. Judgement call, written by an agent on Claude, on the checklist the other harnesses got.

Sculptor 57

Harness reading. Product analytics, error reporting and session recording default to off in the user config, and the app asks for a telemetry level and privacy consent at first run; PostHog runs on the front end only for now (25 of 30). Each workspace is an isolated copy of the code (14 of 20). No prompt-injection guidance (5). No audit log (5). SECURITY.md (8). 25 + 14 + 5 + 5 + 8 = 57.