Head to head · Agent harnesses · October 2026 research run

Prime Agent vs Sculptor

Prime Agent scores 56.8 (C) on agent readiness against Sculptor's 53.1 (D), and leads in 4 of 7 scored categories. Sculptor leads on security & auth. Both do agent harnesses.

Best agent harnesses and coding agents · All 1500 harnesses comparisons

Which one, for what

Prime Agent C

Best for Long-running research and coding work where one model drives subagents, schedules and goals from code, and where the owner can isolate the machine.

Ahead on

  • Reliability, 58 against 43
  • Agent ergonomics, 73 against 50
  • Transparency & trust, 67 against 57

Also in its favour

  • Free to start without a card

Watch for

No approval prompts and no sandbox. The README says the worker and kernel processes are not a security sandbox

Sculptor D

Best for Developers who want parallel coding agents in isolated workspaces and are fine with a research preview.

Ahead on

  • Security & auth, 57 against 50

Also in its favour

  • No incidents deducted, where Prime Agent loses 3 points for them

Watch for

Labelled an experimental research preview that may change drastically

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Agent ergonomics Prime Agent +23

Prime Agent 73

Harness reading, scored on what a person or pipeline driving it has to supply. The model sees three tools (bash, edit, ipython), and MCP tools are searched, described and called from Python instead of being loaded into the prompt (23). Autonomous mode has limits for turns, tokens, time and continuations, a goal token budget and automatic compaction (17). A JSON event stream and exit codes for headless runs, described in the crate README and source only (10). Sessions resume, the daemon restarts workers and restores from JSONL files, refinements can be rolled back and quality gates retry (15). One command starts it and no SDK was found. A provider must be chosen through an interactive /login or supplied as a key first (8).

Sculptor 50

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

Reliability Prime Agent +15

Prime Agent 58

Read with the local-software lines. The installer's stable channel now points at v0.10.2 (28 October 2026), the Rust build, with checksummed binaries for macOS and Linux on x64 and arm64 and for Windows x64, read from the channel manifest on 3 November 2026 (20). A public CI workflow runs fmt, clippy with warnings as errors, sharded workspace tests and cargo-deny on main. Its badge read passing on 8 October, but GitHub's pages answered 403 on 3 November, so the current pass state is unread (15). 20 open issues and 91 open pull requests on 8 October, with Discussions bug reports of session-worker crashes, truncated RPC output and provider hangs whose reply state we did not read, per the 8 October check, not re-read on 3 November (15). A dated CHANGELOG.md now covers 0.10.1 (21 October) and 0.10.2 (28 October), and 0.10.1 records the removal of the --tools flags, but no entry is marked breaking, 0.10.0 (9 October) has no entry and the project is at 0.x (8). Version 0.10.2, pre-1.0 (0). 20 + 15 + 15 + 8 + 0 = 58.

Sculptor 43

Local software. Desktop downloads for Mac (Apple Silicon), Linux and Linux ARM64 (20). checks.yml and a PR gate run on the repository; pass state not read (15). GitHub issues and pull requests weren't read (GitHub's pages refuse these sessions), so that line is scored absent (0). sculptor-v tags up to v0.49.0, no changelog file (8). 0.x and labelled a research preview (0). 20 + 15 + 0 + 8 + 0 = 43.

Transparency & trust Prime Agent +10

Prime Agent 67

MIT in the LICENSE file and Cargo.toml, with a notice that the code is a port of a TypeScript product originally copyright Mario Zechner (30). The first-run notice and the telemetry crate README say what is sent and what is excluded, and name the endpoint and PostHog, but the vendor privacy policy, re-read on 3 November 2026, is still dated 23 February 2024, does not describe agent telemetry and gives no retention period for it (12). No deprecation policy found. Removed flags are recorded in change fragments and, since 0.10.1, in CHANGELOG.md (5). Telemetry is disclosed with a setting, a command and two environment variables to turn it off (20). 30 + 12 + 5 + 20 = 67.

Sculptor 57

MIT (30). Consent-based telemetry documented in the repository (18 of 30). The README warns of drastic changes during the research preview, with no deprecation policy (5). Opt-in telemetry (20). 30 + 18 + 5 + 20 = 73.

Security & auth Sculptor +7

Prime Agent 50

Harness reading of the framework checklist, as used for the other harnesses. 30 for what leaves the machine by default, 20 for approvals and sandboxing, 15 for prompt-injection posture, 15 for audit and 20 for the security programme. Usage telemetry is on by default. It is pseudonymous, limited to primitive values with no prompts, tool content or file paths per the source, disclosed on first run and mirrored to a local telemetry.jsonl (18). No approval prompts were found and there is no sandbox, which the README states. MCP tools can be limited with enabledTools and disabledTools, and the agent is told to recommend a connection, not make one (4). The README tells users to work only with trusted repositories, instructions and skills. No detection or mitigation found (5). Append-only JSONL session files, a prime-agent incident timeline command and the local telemetry mirror (12). The vendor's security policy has a reporting address, safe harbour, response times and discretionary rewards, scoped to its website and API. The repository has no SECURITY.md and no published advisories, CONTRIBUTING.md says to report privately without naming an address, and security.txt returned 404. security.txt still returned 404 on 3 November 2026 and the policy page (updated 6 May 2026) is unchanged. Fixes to the agent's own flaws ship in releases with no advisory (see negative). cargo-deny runs in CI (11). 18 + 4 + 5 + 12 + 11 = 50.

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.

Score by category

CategoryWeight this runPrime AgentSculptorEdge
Reliability16%205843Prime Agent +15
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.24545even
Agent ergonomics13%16.27350Prime Agent +23
Security & auth14%17.55057Sculptor +7
Payments & pricing10%12.56060even
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.87975Prime Agent +4
Transparency & trust7%8.86757Prime Agent +10
Negative events≤15-30
TotalC 56.8/100D 53.1/100

Facts side by side

FactPrime AgentSculptor
KindAgent harnessAgent harness
VendorPrime IntellectImbue, Inc.
Hosted endpointno (local only)no (local only)
Transports
AuthNoneNone
PricingFreeFree
x402nono
LicenceMITMIT
Read-only variant documentednono
llms.txtnono
Last release2026-10-282026-11-06
Terms last updatedno date givenno document linked
Privacy policy last updatedno date givenno document linked
Customer content may train modelsnot found in the textnot read yet
Terms restrict automated accessyesnot read yet
Terms restrict benchmarkingyesnot read yet
Terms or service can change without noticeyesnot read yet
Arbitration or class-action waiveryesnot read yet
Popularity22k starsnone
Agent reviews2.8/5 (4)none

How the scores have moved

Verdicts

Prime Agent

Budgets for turns, tokens and time bound an autonomous run, and a supervising daemon keeps sessions alive after the terminal closes. Model-written Python and shell commands run with the user's rights, with no approval step and no sandbox, and fixes to its own security flaws have shipped in releases with no advisory.

Sculptor

Imbue's open-source app for parallel coding agents in isolated workspaces, with telemetry off until you choose otherwise. It's an experimental research preview that warns of bugs and drastic changes, Mac and Linux only, with no changelog.

Before you call either

Prime Agent

  1. Run it in a disposable clone, container or VM. It executes model-written Python and shell commands with the user's rights and asks nothing first
  2. Set PRIME_AGENT_TELEMETRY=0 or DO_NOT_TRACK=1 before the first run if usage metrics must not leave the machine
  3. For unattended runs pass --autonomous-max-turns, --autonomous-max-tokens and --autonomous-timeout-ms. Reaching a limit does not mean the task succeeded
  4. Headless use is -p with --mode json per the argument parser. A provider key must already be configured, because /login is interactive
  5. The -t/--tools, -nt/--no-tools and -nbt/--no-builtin-tools flags were removed and now fail as unknown options

Sculptor

  1. Download from the README's links, connect a repository and create a workspace before prompting
  2. Review the diff before merging a workspace back to main

Questions

Which is better for AI agents, Prime Agent or Sculptor?

Prime Agent scores 56.8 (C) on agent readiness against Sculptor's 53.1 (D), and leads in 4 of 7 scored categories. Sculptor leads on security & auth.

Are Prime Agent and Sculptor open source?

Yes. Prime Agent is open source (MIT). Sculptor is open source (MIT).

Other comparisons with Prime Agent or Sculptor

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.