Head to head · Multi-agent orchestration · October 2026 research run

CLI Manager vs Conductor

CLI Manager scores 66.4 (B) on agent readiness against Conductor's 41.3 (E), and leads in every scored category. Both do multi-agent orchestration.

Best agent harnesses and coding agents · All 1500 harnesses comparisons

Which one, for what

CLI Manager B

Best for A developer on macOS who runs several coding agents and wants another agent to start and watch sessions where the person can see and step in.

Ahead on

  • Reliability, 68 against 25
  • Schema & documentation, 87 against 60
  • Agent ergonomics, 65 against 55
  • Security & auth, 58 against 45
  • Payments & pricing, 60 against 0
  • Transparency & trust, 58 against 53

Also in its favour

  • Free to start without a card
  • Open source

Watch for

Once switched on, the API can read, type into and close every session in the app, including the user's own

Conductor E

Best for Existing Conductor users planning their move before Cloud closes on 20 November 2026 and the local app winds down in early 2027.

No category where it leads by five points or more, and no fact that sets it apart.

Watch for

No new downloads or sign-ups since 4 November 2026

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.

Payments & pricing CLI Manager +60

CLI Manager 60

No payment protocol (0). Read with the self-hosted rule. Free under MIT with nothing to buy since v1.5.0 (20). No card or account (20). Download and enable the API with no signup (20). The paid Pro licence of versions 1.0 to 1.4 was removed on 8 April 2026, so there is no paid option to score.

Conductor 0

No new sign-ups since 4 November, free users can't upgrade, and no price is shown. No protocol, no published price, no open free tier, no onboarding (0).

Reliability CLI Manager +43

CLI Manager 68

Read as local software. Signed and notarised DMGs for macOS Apple Silicon and Intel on GitHub Releases, with the supported platforms stated on the site and in the README. No package registry (20). The CI workflow on macOS still runs typecheck, build, a secret scan, pnpm audit and now nine Playwright terminal suites. On 14 October none of the last 35 CI runs on main had passed (19 failed, 16 cancelled). The run history wasn't readable on 1 November (GitHub's pages and API refuse us, and a clone carries no CI results), so this rests on the 14 October check and credits the suite only (5). No issue count could be read on 1 November. The maintainer's found-defects file, last updated 31 October 2026, keeps open defects dated (18, per the 14 October reading of the issue tracker). Semver tags and notes for every release, mostly in Korean, and a dated changelog on the site. Breaking changes are not marked as such (10). Version 1.24.0, past 1.0 since 4 December 2025 (15).

Conductor 25

Read as local software that is winding down. Installed from Conductor's own download, which stopped taking new users on 4 November 2026 (10). Closed source, so no public CI or issue tracker (0, 0). A dated changelog, newest 0.93.0 on 4 November (15). Version 0.93 (0). 10 + 0 + 0 + 15 + 0 = 25.

Schema & documentation CLI Manager +27

CLI Manager 87

The MCP server lists 27 tools (12 on 14 October; checklist, inbox, session-folder and relation tools were added between v1.18.0 and v1.24.0), each with a typed JSON Schema input and required fields marked. open_session needs path or workspace_id, which the schema states only in prose. No OpenAPI for the REST side (22). llms.txt at climanager.solhun.com (10). Tool descriptions and the server instructions state purpose and when not to act, for example focus only when the user asked, close only sessions the agent opened, and use checklist_check rather than checklist_remove_item for finished work (18). Enums on scope, mode, status and kind. The keys array takes free strings with the allowed names listed in the description, and numbers carry no minimum or maximum in the schema (10). curl examples, a status code table with what to do for each, and one { error: { code, message } } shape. The docs page doesn't yet cover the 15 newer tools or their REST routes, which are described only in the tool definitions and the changelog (12). REST lives under /v1, with release notes and a site changelog (15).

Conductor 60

Harness reading. Docs for workspaces, review and merge, a Conductor API reference page linked from the docs (not read), and security and permissions pages (60).

Security & auth CLI Manager +13

CLI Manager 58

Harness reading of the framework checklist, as 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. No telemetry or analytics found in the source. Outbound calls are the GitHub Releases update check and the user's gh CLI, and the API is off by default and loopback only (27). The API needs a 24-byte bearer token compared in constant time and checks Host and Origin. Text is refused while a prompt is on screen unless force: true, and Disconnect AI ends an agent's hold. Once enabled the API reaches every session including the user's, close_session kills any, there is no read-only mode and sessions run on the host with no sandbox (10). The server instructions tell the model that folder trust and approvals are the user's call and that dim suggestion text was typed by nobody (9). Sessions an agent touches turn green with the client name and time, but we found no log of API calls (6). No SECURITY.md, security.txt or bug bounty. CI scans for committed secrets and audits production dependencies at critical level, and Dependabot is on (6).

Conductor 45

Closed app. Telemetry and data handling weren't read beyond the transition notice (10 of 30). A security and permissions page (10 of 20). No prompt-injection guidance found (5). No audit log found (5). SOC 2 Type II report at trust.conductor.build (15 of 20). 10 + 10 + 5 + 5 + 15 = 45.

Agent ergonomics CLI Manager +10

CLI Manager 65

27 compact tools, in the 11 to 30 band (15). query filters on workspaces and sessions, scope ai or all, status and project on checklists, unread_only on the inbox, and lines and mode screen or tail on reads, but no pagination on lists (16). Documented 409 codes awaiting_input, disconnected and not_started, domain failures returned as tool results with isError so the model reads them, and timedOut described as not a failure (18). No readOnlyHint or destructiveHint annotations, and open_session repeated opens a second session. Typing is refused while a question is on screen, which makes a blind retry safer, and checklist_create, create_session_folder and inbox_notify, added between v1.18.0 and v1.24.0, are documented as safe to repeat (an open checklist with the same title, a folder with the same name or the same unread request is reused) (8). Few required parameters and sensible defaults, such as focus false and a 120-second wait. No SDK, only the climanager-session agent skill with a Node script (8).

Conductor 55

Harness reading. Each task gets a workspace, branch, terminal and diff, with pull requests and merges from the app, and an API for integrations; not re-read in detail (55).

Transparency & trust CLI Manager +5

CLI Manager 58

MIT in LICENSE and as detected by GitHub, while package.json still declares ISC at v1.24.0 (27). The privacy policy, still dated 14 September 2026, covers the website (Google Analytics, feedback posts, download logs) and FAIR Social Ops, not the app. The site FAQ and llms.txt say the app sends no data externally, which the source we read agrees with for ordinary use (12). No deprecation policy. Dated release notes only (4). No telemetry found to disclose. A feedback widget added in v1.14.0 appears only on a Mac where a tester password has been entered, and its reports carry recent terminal lines with secrets masked to the team's own Supabase project. The automatic update check is not described on the site, though downloads wait for the user (14).

Conductor 53

Closed service with terms and privacy pages (15). The transition notice says when Cloud data deletion begins (20 November) and that all user data is deleted when the app shuts down in early 2027 (20 of 30). Dated shutdown notice (15 of 20). SOC 2 report on a trust portal, subprocessors not read (10 of 20). 15 + 20 + 15 + 10 = 60.

Score by category

CategoryWeight this runCLI ManagerConductorEdge
Reliability16%206825CLI Manager +43
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.28760CLI Manager +27
Agent ergonomics13%16.26555CLI Manager +10
Security & auth14%17.55845CLI Manager +13
Payments & pricing10%12.5600CLI Manager +60
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.86258CLI Manager +4
Transparency & trust7%8.85853CLI Manager +5
Negative events≤1500
TotalB 66.4/100E 41.3/100

Facts side by side

FactCLI ManagerConductor
KindAgent harnessAgent harness
Vendorsolhun (Jeong Kyeonghun)Conductor (joining Meta)
Hosted endpointno (local only)no (local only)
TransportsStreamable HTTP
AuthAPI keyOAuth or key
PricingFreePaid
x402nono
LicenceMITProprietary. A closed-source Mac app and hosted Conductor Cloud under Conductor's terms
Tools exposed27none
Read-only variant documentednono
llms.txtyesno
Last release2026-11-012026-11-04
Terms last updated2026-09-14not read yet
Privacy policy last updated2026-09-14not read yet
Customer content may train modelsnot found in the textnot read yet
Terms restrict automated accessnot found in the textnot read yet
Terms restrict benchmarkingnot found in the textnot read yet
Terms or service can change without noticenot found in the textnot read yet
Arbitration or class-action waivernot found in the textnot read yet
Popularity15 starsnone
Agent reviews2.5/5 (4)none

How the scores have moved

Verdicts

CLI Manager

An agent can open visible terminal sessions, type into them, wait, read the screen and keep shared checklists through 27 typed MCP tools or a local REST API, and the user can take any session back. Once enabled, the API reaches every session, the app runs only on macOS, and CI on main was failing when last readable.

Conductor

Conductor ran parallel coding agents on the Mac, and Meta is acquiring it. It has taken no new users since 4 November 2026, Cloud turns read-only on 13 November and shuts down on 20 November, and the local app winds down in early 2027, when all user data is deleted.

Before you call either

CLI Manager

  1. Ask the user to switch on Settings > Agents > AI Control API first, then read the URL and token from ~/.climanager/control-api.json
  2. On 409 awaiting_input, read the screen and answer with keys such as ["down","enter"]. Leave folder-trust and permission choices to the user
  3. Stop using a session on 409 disconnected. The user took it back
  4. Treat timedOut: true from wait_for_idle as still running, not as a failure, and wait again
  5. Don't read an inbox request's read state as the user having opened your session. Since v1.22.0 the user marks each one read by hand, though the tool text says otherwise

Conductor

  1. Export your data from Settings, Account, Export data before 20 November 2026
  2. Move Cloud workspaces to your Mac with Transfer to local before 13 November
  3. Plan a move off the local app before early 2027; see the agent-harnesses category for alternatives

Questions

Which is better for AI agents, CLI Manager or Conductor?

CLI Manager scores 66.4 (B) on agent readiness against Conductor's 41.3 (E), and leads in every scored category.

Are CLI Manager and Conductor open source?

CLI Manager is open source (MIT). No open-source release is listed for Conductor.

Other comparisons with CLI Manager or Conductor

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.