# Prime Agent (slim)

> Prime Agent is an open-source coding and research agent from Prime Intellect, run from a terminal. The model works through a persistent Python REPL, spawns subagents as function calls and keeps sessions running in a background daemon.

- Full: https://www.anchorterminal.com/tools/prime-agent.md (~7,550 tokens) · this version ~1,530 tokens · JSON https://www.anchorterminal.com/tools/prime-agent.json · canonical https://www.anchorterminal.com/tools/prime-agent
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-08

**C · 60.5/100 · rank #352 of 629 · #9 in Agent harnesses · not agent-ready · confidence medium**


Assessment: Budgets for turns, tokens and time bound an autonomous run, and sessions survive a closed terminal through a supervising daemon. Model-written Python and shell commands run with the user's rights, with no approval step and no sandbox, and the Rust build on the main branch ships only as nightly betas.

## Facts

- Kind: Agent harness · vendor: Prime Intellect · category: Agent harnesses · legal entity: Prime Intellect, Inc. · provenance 61/100
- Auth: None · pricing: Free · x402: no · licence: MIT
- Probe metrics: not measured yet (probes haven't run)
- Interfaces: Terminal UI, `-p` print mode, and `--mode text|json|rpc|acp|daemon` per the argument parser. Daemon commands `agents`, `attach`, `status`, `doctor`, `schedule`, `shutdown`
- Built-in tools: `bash`, `edit` and `ipython`. Subagents, compaction, goals, skills and MCP calls are Python functions in the persistent REPL
- Approvals: None found. Commands and edits run with the user's rights. MCP tools can be limited with `enabledTools` and `disabledTools`
- Sandbox: None. The README recommends an external sandbox or restricted environment for untrusted code or instructions
- Network: No egress controls found. `--offline` exists as a flag
- MCP client: stdio and HTTP servers through `prime-agent mcp add|list|get|remove`, OAuth through `/mcp login`, and a built-in service catalogue
- Models: `/login` for a subscription or an API-key provider. Provider code for Anthropic, OpenAI, Google, OpenRouter, Mistral, Bedrock, Azure OpenAI and Prime Inference
- Self-improvement: `/refine` applies small updates to supplemental prompts, memories, skill descriptions and subagent specifications, with snapshots for rollback. The base system prompt is not rewritten
- Telemetry: On by default, pseudonymous installation id, primitive values only, sent to api.primeintellect.ai and forwarded to PostHog. Off with `/telemetry off`, `PRIME_AGENT_TELEMETRY=0`, `DO_NOT_TRACK=1` or offline mode
- Release channels: Stable v0.9.8 of 29 September 2026 (TypeScript build). Nightly v0.9.9-beta.55 of 7 October 2026 (Rust build). Windows on the beta channel only
- Platforms: macOS and Linux on x64 and arm64 as checksummed binaries, with an npm install needing Node.js 20.6 or later as the fallback
- First public release: Announced on 5 August 2026 (https://www.primeintellect.ai/blog/prime-agent)

- Scores: Reliability 65, Performance pending, Schema & documentation 42, Agent ergonomics 73, Security & auth 50, Payments & pricing 60, Task success pending, Maintenance & community 79, Transparency & trust 64 · total over the 7 assessed categories
- Why: Reliability, Read with the local-software lines. · Schema & documentation, Harness reading of the framework line. · Agent ergonomics, Harness reading, scored on what a person or pipeline driving it has to supply. · Security & auth, Harness reading of the framework checklist, as used for the other harnesses. · Payments & pricing, Self-hosted rule. · Maintenance & community, Nightly v0.9.9-beta.55 on 7 October 2026 and stable v0.9.8 on 29 September 2026 (30). · Transparency & trust, 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…
- Sources: 28, open questions: 9, both in the full twin
- Capabilities: agent.harness, agent.mcp-client, agent.multi-agent
- JSON: https://www.anchorterminal.com/api/v1/tools/prime-agent.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/prime-agent.svg` or a link to https://www.anchorterminal.com/tools/prime-agent from a page on primeintellect.ai or one of its subdomains, or the README of github.com/PrimeIntellect-ai/prime-agent, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

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

## Connect

```bash
curl -fsSL https://app.primeintellect.ai/prime-agent/install.sh | sh
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| goose | BB | 73.9 | agent.harness, agent.mcp-client, agent.multi-agent | https://www.anchorterminal.com/tools/goose.min.md |
| Qwen Code | BB | 72.4 | agent.harness, agent.mcp-client, agent.multi-agent | https://www.anchorterminal.com/tools/qwen-code.min.md |
| Gemini CLI | BB | 72 | agent.harness, agent.mcp-client, agent.multi-agent | https://www.anchorterminal.com/tools/gemini-cli.min.md |
| OpenHands | BB | 70.8 | agent.harness, agent.mcp-client, agent.multi-agent | https://www.anchorterminal.com/tools/openhands.min.md |
| OpenCode | B | 67.7 | agent.harness, agent.mcp-client, agent.multi-agent | https://www.anchorterminal.com/tools/opencode.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)