sandbase-harness

Source#

• Repository: sandbaseai/sandbase-harness
• Last source update: 2026-08-16
• Last verified: 2026-08-16

Tags#

Integration notes#

Repository is focused on tool/server interoperability; wrap in bounded worker contracts for production use.

worker.md example#

Starter worker.md contract mapped from this registry entry. Copy this file and adapt schemas, constraints, and statuses for your task.

---
id: sandbase-harness-repo-derived-worker
name: sandbase-harness Repo-Derived Worker
version: 1.0.0
source_registry_url: https://worker.md/registry/sandbase-harness/
source_repository: https://github.com/sandbaseai/sandbase-harness
repository_default_branch: main
repository_language: TypeScript
repository_license: Apache-2.0
repository_updated_at: 2026-08-16
worker_mode: tool-gateway-worker
derivation_method: github_repository_metadata_plus_raw_readme
derivation_confidence: 0.95
derived_on: 2026-08-16
tags:
  - agent-framework
  - agent-observability
  - agent-runtime
  - agent-sandbox
  - ai-agents
  - ai-infrastructure
---

# sandbase-harness Repo-Derived Worker

## Repo-derived summary
- Registry summary: Open-source CMA-compatible agent runtime for any model, with MCP tools, sandboxed sessions, audit, replay, and a local console. Includes a native DeepSeek Harness bundle over stdio MCP.
- Repository description: Open-source CMA-compatible agent runtime for any model, with MCP tools, sandboxed sessions, audit, replay, and a local console. Includes a native DeepSeek Harness bundle over stdio MCP.
- Stars (snapshot): 593
- Primary language: TypeScript
- Worker mode classification: tool-gateway-worker

## Extracted from
- https://github.com/sandbaseai/sandbase-harness
- https://github.com/sandbaseai/sandbase-harness/blob/main/README.md
- https://img.shields.io/github/stars/sandbaseai/sandbase-harness?style=social
- https://img.shields.io/github/v/release/sandbaseai/sandbase-harness
- https://img.shields.io/github/discussions/sandbaseai/sandbase-harness

## Evidence notes (from repository text)
- README summary paragraph: A local-first runtime for AI agents. Sessions, sandboxed tools, memory, credentials, audit trails, and a built-in Console — all running on your machine or in your own infrastructure.
- A local-first runtime for AI agents. Sessions, sandboxed tools, memory,
- | Inspect long-running agents | Persistent sessions, resumable event streams, audit, and replay |
- | Control tool access | MCP toolsets, credential vaults, permission policies, and approvals |
- Agent SDKs handle the model loop. Production agents need more: persistent
- sessions, tool governance, sandbox boundaries, credential handling, memory,

## Installation hints found in README
- `npm install managed-agents`.`

## worker.md contract (derived starter)
Purpose: Expose repository-supported tool/server capabilities behind a bounded worker interface.

### Input schema
```json
{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "request_id",
    "operation",
    "payload"
  ],
  "properties": {
    "request_id": {
      "type": "string"
    },
    "operation": {
      "type": "string"
    },
    "payload": {
      "type": "object"
    }
  }
}
```

### Output schema
```json
{
  "type": "object",
  "additionalProperties": false,
  "required": [
    "request_id",
    "status",
    "result"
  ],
  "properties": {
    "request_id": {
      "type": "string"
    },
    "status": {
      "type": "string",
      "enum": [
        "ok",
        "retryable_error",
        "invalid_request",
        "invalid_output"
      ]
    },
    "result": {
      "type": "object"
    }
  }
}
```

### Constraints
- timeout_seconds: 30
- max_attempts: 2
- idempotency_key: request_id
- status_enum: [ok, retryable_error, invalid_request, invalid_output]
- notes: adapt to concrete APIs/classes documented in this repository before production use

## How this should be used
1. Treat this file as a repo-derived starter profile, not a claim of an official repository API contract.
2. Replace schemas with exact interfaces from code/docs you adopt.
3. Keep execution bounded and auditable using worker protocol constraints.

How to use#

• Save this as a worker spec file (for example: sandbase-harness-my-task.worker.md).
• Replace the input/output schemas and purpose with your real bounded task.
• Enforce schema validation + timeout + retry policy in your runtime before production use.

Citation#

Reference URL: https://worker.md/registry/sandbase-harness/

Source URL: https://github.com/sandbaseai/sandbase-harness