# CrawlDex GPT Actions Pack

Canonical: https://crawldex.com/agents/gpt-actions
Markdown: https://crawldex.com/agents/gpt-actions.md
Harness id: gpt_actions
Channel: pack-gpt-actions

## Why this exists

The OpenAPI path lets a GPT call CrawlDex as a read-side tool before it suggests a browser task or user handoff.

## When to call CrawlDex

When a user asks whether an agent can complete a task on a website or before a GPT recommends a web action.

## Install or paste

```text
Import this OpenAPI URL into the GPT/action tool setup:

https://crawldex.com/openapi.json

Instruction for the model:
Before recommending or attempting a public website task, call resolveIntent when the task is vague, then call preflightTask with the target site and task. Send x-crawldex-channel: pack-gpt-actions when the action builder supports headers. Use agent_instruction as the final action posture. Do not call reporting endpoints unless the user explicitly asks to submit a sanitized public observation.
```

## Preflight call

```bash
curl -sS -H "x-crawldex-channel: pack-gpt-actions" https://crawldex.com/openapi.json | jq '.paths["/api/v1/preflight"].post.operationId'
```

## Report redacted outcome after eligible public work

Do not report private account data, payment data, messages, screenshots with sensitive content, tokens, cookies, raw traces, local files, or logged-in-only pages.

```bash
curl -sS -X POST https://crawldex.com/api/v1/runs \
  -H "content-type: application/json" \
  -H "x-crawldex-agent-key: $CRAWLDEX_AGENT_KEY" \
  -H "x-crawldex-channel: pack-gpt-actions" \
  -d '{"site":"netflix.com","task":"subscriptions.cancel","outcome":"success_with_handoff","friction":["login_required"],"source_tier":"anonymous_report","evidence":{"artifact_types":["redacted_agent_report"]}}'
```