---
title: Logs
description: "Logging and monitoring"
---

{/* Auto-generated from OpenAPI spec. Edit the overview at api-overviews/logs.mdx, not this file. */}

The Logs API returns **individual tool execution events**, one record per tool call. Use it to debug failures, inspect request/response payloads, and trace specific user activity. For aggregated counts (how many tool calls happened), use the [Usage API](/reference/api-reference/organization) instead.

All endpoints in this section require a **project API key** (`x-api-key`) or a valid session cookie.

## List logs

```bash
curl -X POST https://backend.composio.dev/api/v3.1/logs/tool_execution \
  -H "x-api-key: YOUR_PROJECT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "limit": 20,
    "time_range": {
      "from": 1744848000000,
      "to": 1744934400000
    },
    "filters": [
      { "field": "toolkit_slug", "operator": "==", "value": "gmail" },
      { "field": "status", "operator": "==", "value": "failed" }
    ]
  }'
```

The response contains a page of log entries and a `next_cursor`:

```json
{
  "logs": [
    {
      "id": "log_-jRTWClpBoVo",
      "timestamp": "2026-04-17T10:25:00.000Z",
      "type": "tool.execution",
      "status": "failed",
      "level": "error",
      "message": "GMAIL_SEND_EMAIL failed: invalid recipient",
      "metadata": { /* tool, toolkit, user_id, connected_account_id, ... */ },
      "metrics": { "duration_ms": 202 },
      "parent": null
    }
  ],
  "next_cursor": "eyJwYWdlIjoyfQ=="
}
```

Pass `next_cursor` back as `cursor` on the next request to paginate. When `next_cursor` is `null`, you've reached the end.

### Filter fields

Pass one or more filters in the `filters` array. Filters are **AND**-combined.

| Field | What it matches |
|---|---|
| `tool_slug` | The specific tool that was called (e.g. `GMAIL_SEND_EMAIL`) |
| `toolkit_slug` | The toolkit (e.g. `gmail`, `slack`, `github`) |
| `connected_account_id` | The connected account used for the call |
| `auth_config_id` | The auth config (integration) behind the connected account |
| `status` | `success` or `failed` |
| `user_id` | Entity that initiated the call |
| `session_id` | Tool router session, if routed through a session |
| `sandbox_id` | Sandbox the call ran in, if applicable |
| `request_id` | Request ID (useful for correlating with your own logs) |
| `log_id` | Exact log ID (equivalent to the detail endpoint) |

### Operators

| Operator | Meaning |
|---|---|
| `==` | Exact match |
| `!=` | Not equal |
| `contains` | Substring match |
| `not_contains` | Substring does not match |

### Parameters

| Field | Type | Default | Notes |
|---|---|---|---|
| `limit` | number | `20` | Max 100 |
| `cursor` | string \| null | `null` | Opaque pagination token from previous response |
| `filters` | array | `[]` | AND-combined |
| `time_range.from` | number | — | Epoch milliseconds |
| `time_range.to` | number | — | Epoch milliseconds |

## Get a single log

Fetch one log by ID to get the **full** payload, including request/response bodies, timing breakdowns, and source metadata:

```bash
curl https://backend.composio.dev/api/v3.1/logs/tool_execution/log_-jRTWClpBoVo \
  -H "x-api-key: YOUR_PROJECT_API_KEY"
```

The detail response includes everything from the list shape plus:

- `timings`: `start_time` and `end_time` in epoch ms
- `context`: `session_id`, `trace_id`, `request_id`
- `source`: `host` (e.g. `mcp`, `sdk`, `api`), `framework`, `language`
- `data`: the full request payload and response body

This is the endpoint to call when you need to reconstruct *exactly* what happened, for example when debugging a 500 from a user report.

## Recipes

### Find failed Gmail tool calls in the last hour

```bash
NOW=$(date +%s)000
HOUR_AGO=$(( $(date +%s) - 3600 ))000
curl -X POST https://backend.composio.dev/api/v3.1/logs/tool_execution \
  -H "x-api-key: YOUR_PROJECT_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{
    \"time_range\": { \"from\": ${HOUR_AGO}, \"to\": ${NOW} },
    \"filters\": [
      { \"field\": \"toolkit_slug\", \"operator\": \"==\", \"value\": \"gmail\" },
      { \"field\": \"status\", \"operator\": \"==\", \"value\": \"failed\" }
    ]
  }"
```

### Get failures for a specific user

```bash
curl -X POST https://backend.composio.dev/api/v3.1/logs/tool_execution \
  -H "x-api-key: YOUR_PROJECT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "filters": [
      { "field": "user_id", "operator": "==", "value": "user_abc123" },
      { "field": "status", "operator": "==", "value": "failed" }
    ]
  }'
```

### Fetch a single log's full request/response

```bash
curl https://backend.composio.dev/api/v3.1/logs/tool_execution/log_-jRTWClpBoVo \
  -H "x-api-key: YOUR_PROJECT_API_KEY"
```

## Endpoints

<ApiEndpointsTable endpoints={[{"method":"POST","pathV31":"/api/v3.1/logs/tool_execution","pathV3":"/api/v3/logs/tool_execution","summary":"Search and retrieve tool execution logs","href":"/reference/api-reference/logs/postLogsToolExecution"},{"method":"GET","pathV31":"/api/v3.1/logs/tool_execution/{id}","pathV3":"/api/v3/logs/tool_execution/{id}","summary":"Get log details by ID","href":"/reference/api-reference/logs/getLogsToolExecutionById"}]} />